Video 1: Choosing Research Question

In this video I'll teach you how to choose a good research question for your physics IA.
Having a good research question doesn't mean you must get level 7. However, if you get a
bad research question to start with, then it would be the same as starting with your wrong
foot. So in that case you may not even get level six or seven, no matter how hard you try in
the later stage. So choosing a good research question is very very important.
Quality of a good research question. So let's talk about what makes a good research
question. First of all, like what we said in the previous video, you should only do one
independent variable because all of the 12 pages you don't have that many content or space
to talk about two independent variables. So just do one and make sure that it's a very
focused research question.

Secondly, you should probably think of some background research and then after you find
that, it will be nice if you can simply study the same thing but then you're using a different
methodology. So for example, if you are studying on the magnetic force, and there may be a
different way to measure the magnetic force in your setup. So that could be something. It
would be appreciated if you can try to think often something neat and maybe different,
slightly different from the existing research. The best thing is in that case you can then
compare your method with other people's method.
Number three is that your research questions should have a strong physics background and
or framework to support. They should say hopefully before you start you should already find
there's a certain formula or there should be some research done by other people. So that
you can follow and probably you can already see the independent and dependent variable
has a certain mathematical relationship. So for example it could be like a straight line or
quadratic equation, exponential equation, etc because only with such a strong framework uh
that in that case then you could have a very clear mind and direction when you do the
evaluation and analysis. Even though from the official guideline from IBO said oh it is not
really necessary you can still where there is no existing framework supporting or explaining
your research question you can still kind of using the common physics knowledge you
guess.
But I would say uh within the usual physics IA remember this is not EE, this is just IA. So the
topic you do very likely it has been done by many other peoples already so there should be a
very well established relationship that you could find. If you cannot find it just because you
didn't try hard enough.
And lastly uh it may be also good if you can I mean uh to to make it more scientifically
interesting if you want to get uh to two months a high spend in the personal engagement.
One way to one type of really special good research question is to find the constant that we
learn in physics. So it could be well I think for gravitational constant that may be too hard but
if you can try to look for things like the say the gas constant, Coulomb constant, permeability
of free space. Those are something that you could work on and it could be quite interesting
too. But of course that shouldn't be the only thing that uh you should do for research
question are there actually many people who are not finding constant just finding relationship
between x and y however it should be so surely relevant so that you should say like we said
in the previous video if you try to find the refractive index for something like uh the syrup how
that will affect by the temperature it doesn't sound like socially relevant because why do you
like care about that in the society or in your daily life and it probably should not be in the IB
service directly.
Okay, I'll explain this part later on. Quality of a bad research question. Bad research
question, so the first example is saying it cannot be repeated by other people easily and in
fact this is not just about your research question but also about the methodology that you do
so. For example, if your experiment involved too much uh human as in maybe in your
methodology it relies on how you drop something or how the environment it may be unique
then it may not be a good method or even not a good research question.
So I had students in the past are suggesting studying topics like running or cycling. So those
are topics that if you cannot think of a methodology that you can reduce the involvement of
human uh that it won't be a good research question or methodology. Secondly, comparing
the brands of different products, so this is the same as biology and chemistry as well and it
has been explicitly mentioned in somewhere in the exam report I've seen before. So same
for physics so for example if you're comparing different brands of soccer shoes for example
is actually not a really good research question because again first of all your independent
variable should be quantifiable. So if you just do different brand ABCD it's not something that
you can describe with numbers. And ultimately you should link your two variable
independent and dependent variable together with the physical physics framework or
concepts. So having a different brand it is actually quite hard to tell uh the design or the like
whatever internally it has been made for whatever product that you are studying.
So uh just try to simplify your setup and choose a quantifiable independent variable instead
of comparing the brands. Number three like we said actually earlier if there's no established
relationship we can find at the same time you cannot derive anything from the physics law
then probably you should not do it. There are times where you could still derive something
for example if you're doing mechanics quite likely you will start off with Newton's law so for
example like f equals m a and probably you can derive something and maybe together with
the free body diagram also. So if that's the case you can derive some interesting formula
from the basic law in physics then it will be nice right and even if you cannot really find much
about the research from other people is still acceptable but of course like I said earlier it
would be the best if you can do both of these.
Number four there's actually something called a list of prescribed lab in IB physics which
your teacher are expected to cover these experiments with you so these are like the very
very basic practical activity that I be kind of assigned to all the teachers that you have
recovered and through these experiments you can learn the basic technique for doing
experiment. So you can see a list here and you can try to take a look so say if you are
having your research question in determining the acceleration of free-fall that may not be a
good idea if you want to achieve a high higher level or doing maybe finding specific latent
specific heat capacity of water using the kilometer check techniques again it is not
something good or internal resistance etc.
You can try to obtain insights from here all right so lastly then you may ask hey Mr Wong can
you give me some hints then how can I think of a good like research question can you just
tell me one. My answer is that uh of course you can go and google it but those you find will
be quite commonly done and I would say it may be the best of your teacher's interest and
also the market's interest to show something new I would say so show something that is
more more unique and really you don't just find a question online but then you find the
relevant concept that you find interested first. And then you develop it into a research
question.
So if you google for quite a bit you will find there is a document called like 300 stimulating
ideas for IB like physics idea and you will see a lot of different ideas. Uh if you like to read
that's fine but I I would say I'm not all of these will be good for you like for choosing a physics
IA. Alternatively, I would like to recommend you to go for a website called HyperPhysics. Not
only useful for your IA ideas but it's also very useful for your revision in the future as well.
So this is a website of hyperphysics and you can see it's shown in a mind map that's why I
said it is very useful for revision in the future. For talking about the IA ideas you can click into
whatever topics that you find interested in say mechanics and say you can see in like inside
mechanics there are so so many other sub topics that you can look for. So things like maybe
our conservation of angular momentum could be one of the thing that you you can study.
And say inside you can take a look of how we actually measure angular momentum and like
I said uh in the first place in this video you need a good framework established framework
that is probably a formula that governs your research question.
So like this one it will be very good to show you uh the relationship that has been fined by
you know the people in the past already and so your IA can hopefully use a maybe an
interesting method to verify this and the way that you can choose is obviously you need the
independent variable and dependent variable. So out of these formula maybe you can
choose one of them to be the independent variable and the other one to be the dependent
variable. And so the rest of the equation okay there may be more than three right so the rest
of the variables should be constant that means the control variable from your whole
experiment.
So that's why if you try to look at this website hyperphysics it will be very helpful because it
would just eventually tell you some formula at the end. All right so this will be something very
useful direct and help you to formulate your ideas number two. This is actually the way that I
find uh usually the best IA idea and I encounter will be using this approach. So you can just
go and review take a look of all the physics concepts that you learned in the past. So not
only IB physics so it could be also the IGCSE physics as well and think try to take a look and
see which concept that you learn but you only learn in the qualitative way.

"okay so that is you say oh you may just say uh when a increase b would decrease
something like this right but then it never mentioned whether it will increase or decrease or
whether or not it has whatever relationship is there inverse relationship or uh what what kind
of relationship that uh it actually be quantitatively so those questions actually are very good
very good question that also coincide with the idea of scientifically interesting in the previous
slide so example could be like how the rate of evaporation affected uh by you know different
factors that you learn in igcse or it could be we all know that away from the madness the
magnetic field strength is actually weaker but then uh how does it decrease like what kind of
mathematical model would be suitable to describe is it just linear or would it be inverse
square or what another example could be like we all know a resistance even ib as well it will
increase with the greater surface area so how how does it actually increase does it again
increase linearly or does it increase in what kind of ratio so actually all these uh there are
actually formula that of course are found and established by the previous researcher and
you can always try how to google it and find it out and again it will be very very interesting
and topics for you to do for your ia and there are so many different concepts you can talk
about within each topic so really go and take a look and review all the concepts that you
learned in the past number three this is something that i would recommend for my ee
student i think it's more suitable for ee but if you are really strong in physics but you you're
taking some other extended essay i guess this is fine for you to think about this as well so i
ypt is actually the name of international young physicist tournament and it's just a
competition that is a worldwide and even uh some of you if you're again really strong you
may want to participate in this but what i want to focus on today here is uh you can see there
are different problems not only this year but also in the maybe past two years and you can
see there actually quite some research questions that are extremely extremely challenging
and interesting so things like uh i just randomly uh read it really like here it said uh tesla
valve so it also tells you some new ideas as well actually i have no idea what tesla wealth is
but then if you read through all the questions you should be able to find at least one or two
that is doable and also interesting so like here you can see it's just a valve that has a certain
geometry shape passive one direction valve and basically this is the whole research
question already so try to study create a certain valve this valve and study the relevant
parameters which you can of course research on the internet and so the best thing is i think
uh all these questions again um i would say if you take the question from here you should
really give credit to this competition and this website as well because you are not the one
that come up with such a problem uh but again i would really appreciate uh the quality of the
problems that have been posed here and i think the best part is uh since they this
tournament itself is for secondary school student anyway so it should be of a very suitable
difficulty for you to solve as well assuming that you are very passionate and strong in
physics number four 3d printing technology so assuming you have a 3d printing at home or
you have one at school maybe from your dt subjects then this is something that you can
think about as well so in the past uh the physics ia may be limited by the resources that you
have in the lab uh or it's just simply hard for you to put in the resources to find you know
spend so much time to find a certain material so having know how to use 3d printer and
design certain things can enable you to study some topics that is that was very hard to study
in the past in a normal high school setting so for example in the past if you want to design a
certain experiment that would need different size of say a tube or cylinder it may not be easy
to find all the size in the hardware workshop but then if you could simply design it with your
cad software and print it out it would just be quite straightforward and simple and then you
can just do whatever experiment you want to do you can actually also go to a website called
thingiverse where you can download free free free 3d printing files uh and you can actually
change the parameters on them so say you can also say this one you can then simply study
uh how the length will affect the frequency of course this i don't think is a really good
research question anyway but i think this is just to show you an idea after you get a certain
file you can actually further change them manually in the cat software if you know how to do
so so this will be something that will get handy and also making your ia very unique as well
and lastly if you still don't have any ideas it may be good if you just take a quick look on what
kind of apparatus you have at school or what kind of vernee equipment the data logger you
have in school because maybe there are sometimes you want to do a certain experiment
you just don't have the right uh data log at school so just simply looking at what you have
available may help you to think of what to do directly that's all for this video if you like this
video please comment subscribe and like in the next video we'll start to talk about how to
actually write your id for each part you"

Video 2: Introduction and Framework

"Let's begin introduction and framework. First of all, for introduction, let me just reiterate
once again, do not think this is a session for scoring personal engagement, and this is
something that you have to show throughout the whole IA report, right? So it's not just about
introduction. So saying something like, 'When I was a child, I'm fascinated with blah blah
blah, fascinated with cars, or etc.' would not be meaningful, right? So we try to say
something that is fully relevant to the physics concept that you are studying and also
something else. So, what exactly do you need to talk about will be here. Let's take a look.
So, you will want to show something that is describing and explaining why this is
scientifically interesting. So what it means, it could be, how is this related to the daily life? It
could be not just your daily life, it could be general people's daily life, okay? It could also be,
how do we have more advantages if we understand this? So for example, when you study
the transformer, maybe there is some factors that you can study so that, in that case, when
we try to convert the voltage from one to another, it will be the most efficient. Or it could be
studying some other things like how the safety cushion, right, the volume of it will be the
most effective. Or it could be something like how a certain energy resources for a certain
factor, it could generate the most energy. Other than that, you should also talk about past
researcher. So, this part is a little bit similar to what you do for a literature review in EE as
well. What you could talk about is, you may try to state also in maybe in the past there is a
researcher doing a certain methodology and somehow you've thought of another
methodology that you want to try and compare method. Or it could be a different range of
independent variable. So maybe he or she was studying a certain range of temperature
which was only range from say 20 degrees Celsius to say 60 degrees Celsius, and you want
to expand that to maybe up to 100 degrees Celsius or go lower up to say negative 10
degrees Celsius. And that of course will be very helpful in pushing our understanding
towards the world. Or it could also be if you say, all your final researches ABC and maybe
they have a little bit difference in terms of the result, maybe some of them final equation is
different from the others, or maybe it's just the coefficient like, you know, maybe in some
equation it could be like a equal to K X square, right, or something like that, right? Not a
actually, should be Y, right, Y equal to K X square, and maybe that K is just simply different
from different researchers. So maybe you want to find out how that would be affected. When
you talk about those researchers, you may also want to show more in-depth information,
especially about the research finding, and that could be show more diversity by showing
their graphs. So whatever independent dependent variable that they are doing using the
graph, then you can show what they find out basically. So maybe it's a linear relationship, it
could be a curve or whatever trend that they find out. So telling us directly will be very useful.
If you can't find a graph or even the data table, well actually if you can find a graph, you can
introduce that data table to plot a graph by yourself. Worst case, if you don't have anything,
at least you should be able to find the formula that they suggest in the research paper.

For framework, I would say this is like different from introduction. So introduction you can
talk about other people and also how that research question is going to be beneficial to our
world or to yourself at least. For framework, it's more focusing on the concepts that is related
to wrestling to your research question. So what you should do is surely 100 percent you
should derive the relevant formula from the definition in physics. So it could be like maybe
you are doing mechanics, there may be F equals to Ma, or maybe you are doing circuits,
maybe you're relating to the power, that may be power equals to IV or anything, or voltage
equal to energy over charge, something like that, right? That by definition. And then in that
case, hopefully the best scenario is you can express the relationship between your
independent variable and dependent variable, which should be seen in your research
question territory. If you really don't know how to derive it, and but if you could find the
relevant formula from other people's research, try to see how they prove it, usually they do,
alright? Or at least try to search that formula on Google, and pretty sure there is a certain
way that people prove it. In a
 case where it is really coming out from the experiment, then you should try to find more
sources and see whether there is any difference from each other researchers. The other
thing that is also very important and people overlook as well is, you should also state the
required assumptions here, which is needed to support the derived work that you show
above. For example, you may be using some kinematics equation that we use, what we
actually learn in IB, right? So it's quite simple, but then you still need to clearly describe and
state those assumptions because we don't expect you to to know like everything in IB. So
you have to reiterate what assumption that is required. And there is very useful later on,
maybe you'll find out throughout your whole experiment, those assumptions generally was
not met. So that's why maybe your result is different from the theoretical one. So for
example, maybe if you're using kinematics equation, say V square equal to U squared plus
2AS or other kinematics equation, then you should remember that the assumption was,
acceleration has to be a constant, right? But maybe probably in the actual situation in your
practical activity, acceleration cannot be constant simply. So that kinematic equation actually
made a false assumption. Or you can evaluate this at the end of your IA report. Or maybe
you're, if you're doing something related to magnetic field, then maybe one of the
assumption is there's no external magnetic field. Maybe you have a magnet or coil to
produce the magnetic field in your experiment, then maybe the assumption is there's no
external, that is probably the earth magnetic was not considered in the framework. So all
these assumptions has to be stated very, very clearly in a list. Next thing, which again I find it
would be extremely useful to show a good personal engagement and also exploration, that
is, if you could discuss the relevant physics concept in your research question in the
macroscopic view, that is a particle view, that would be nice. I think most of the topics
actually you could actually do microscopic view. So especially for maybe thermal and also
entropic or even waves as well, you should be able to, like, talk more about the microscopic
view. So not just using the formula, play around with the mass, and I rearrange them, but
also you could explain microscopically in a qualitative way. And lastly here, which something
I don't recommend, is that you'll make a false claim or use some very strange way of
guessing, right, without any citation or scientific support. Some of you may form a thing of
doing hypothesis, and in fact, in the document from IBO, they said hypothesis actually is not
needed, right, it's really just optional. So I personally say I don't think making a hypothesis is
too meaningful anyway. I would just say I rather you keep focusing on what you can derive in
theory from the fundamental definition or from other people's research, and that will be pretty
good enough. Because quite often when you try to make a hypothesis, then you kind of
make a false claim, or you usually try to use your intuition to protect a certain result, which
will not be appreciated."

Video 3: Pilot Study

"in this session we'll talk about pilot study before we start I really have to say if you go on
YouTube and search about like all sorts of video talking about IA you find out pilot study this
session has been much much underrated many people didn't even include this part and I'll
explain to you why this is important and what you must include this in your IA if you want to
score high before we talk about how we write it let's talk about the reason why we need to do
pilot study first so first of all of course we want to score better and that is referring to the
eternal engagement because in this case you can really show to the marker that why you do
these and do that and design your experiment later on and then we'll also effects the
methodology which means the respiration will also be improved as well more practically
speaking when you do a pilot study that means before you actually do your final investigation
you also want to see whether your result will be obvious and that is to say within your
independent variable range for example well maybe you're doing how temperature affects
some things or maybe you can do 20 degree Celsius to 100 degrees Celsius at the end but
then during your first study maybe you want to pick simply 20 degree and 100 degree and
see whether or not it will give a good result for dependent variable maybe just maybe when
you do the experiment and your final hate with these two extreme you also get maybe only
5% of change in your dependent variable then maybe this is not a good methodology for you
to try or not a good range for you to try so maybe there's something you want to change in
that case you can either change the range you can either change the methodology or even
you can think about changing the control variable so that may be the dependent River will be
having a greater range so this is something that we do and also something to consider
before you do the actual investigation another example that I can use to help explain why the
policy is important could be a research question related to tapas effect so in case you are in
SL you may not know this this is about how a moving source of sound can emit the Sun with
a frequency that is appear to be higher or lower so in the experiment the student probably
having a ball which has a buzzer inside emitting a sound and they want to change the
frequency and see how the apparent frequency will change but the thing is because of the
speed that they rank the reach is not high enough so maybe at the end they find out or even
by calculation the frequency that you could actually change is just probably three percent
only so it's really not enough for you to calculate anything or measure anything because
probably at the end the change of frequency is probably only like five Hertz only right so it's
not actually measurable because the uncertainty of finding the frequency is probably the one
not to hers already so one way of changing it if you could somehow manipulate the controller
or because in this experiment the original frequency is a constant right it's one of the control
variable so if you can actually maybe multiply these frequencies or change the frequency 10
times larger and of course assuming that is still within the audible range that is measurable
by really quick and then obviously even though it is still 3% with the same speed limit then
the Delta F probably will also be 10 times graders will be like 50 Hertz right for the change so
probably it's more measurable in terms of the absolute value so this is an example to show
you how you can change the control variable in your research question so that your
measurement can be more reliable and probably not affected by the random error as well so
really pilot study is a way to help you to get evidence so because we try it right you may try
like randomly or a chopper trial-and-error of different control variable and then you may find
out for maybe a certain setup you can get the public the best range of dependent variable
and this is a way to justify your choices of the range of independent label and also the
setting of control rel so to summarize what to write under “Pilot study”? we'll be talking about
the those trial runs that you do and for each trial run you may want to talk about the choice of
contourable because you may have changed the control variable for each trial right so then
you can see which one can give you a more office result and also you will also of course
describe the range of the independent variable as well and see whether you can extend it
even more sometimes it may not be related to the controller itself in may like a more major
issue could be methodology so it I would just say stable problems that you face you find here
so it may not just be control but it could be really something else that you may not be able to
solve directly so maybe your way you could change in methodology so then you can do
something more meaningful and the data is also measurable and suitable for analyzed later
and there are also cases where some of the student that I encounter in the past that they
have planned so well they try to execute the procedure and collect data and even by the
time where they collected data they try to plot a graph eventually they find out there is some
much clear trend or maybe the random error is too large and at the end they kind of get to
redo the whole experiment again in that case you can also put all these effort that you have
spent on the previous study into this so-called pilot study alright so pasady is something that
you can talk about something maybe failed at the first place and how you can make a
change in the design of the final investigation at the end so again the whole purpose is to
justify why you make the final design and also hopefully to impress the market that you really
personally engage into this topic so make sure you try to explain all the consideration as
much as possible"

Video 4 : How to write Variables

"in this video we'll talk about the variable session in this part you need to talk about the
independent variable, dependent variable, and also the control variables. I will always
recommend my student to do a table to represent it because it's so much easier to follow in
this case and also you also make sure yourself put on all the relevant information. So you
can make a table like this and also for the variable you can put down the symbol as well. So
like I said if say whenever you put down a certain variable say temperature then you can put
T next to it maybe you are trying to find out the density for example for how maybe
temperature at the densities or density will be rho in that case for control variable if you don't
want to that's fine if you don't put symbol to it but then I would say at least you should think
of three relevant control variable and in fact the more the better of course you have to put
something that is relevant alright so not just say something that is too obvious like all the
same guy operating the experiment shouldn't be put here. So inspiration you can think from
the framework because in the earlier session you have deduced the formula or from other
literature's you'll find a formula so from the formula you should be able to see many more
other parameters and those very likely should be the control variable for the description let
me show you in the next slide. So for independent variable simply is just to put on the range
of what you do and also maybe the interval as well if you want to so let me just put that
interval here so let's say for the temperature you may want to do say like I said always say
20 degree to say 100 degree or maybe even lower you can actually do lower than that so
maybe zero degrees Celsius and maybe the interval is actually by 10 degrees so in fact you
can't list them out so zero degree 10 20 30 etc all right so then people can see the exact
interval that you are doing cuz that will be much much easier to understand later on when
people try to understand and read your procedures and of course more importantly you
should put on a justification why do you do this why don't you do something else and so this
part should coincide with what you put down earlier in the pilot study. Next thing about the
dependent variable should be how it is defined it may be coming from the process data so
like the example that I gave in the previous slide if I want to find density then of course it was
defined as mass over volume so that means if you try to explain how is measure you have to
talk about also I will measure the mass first and then measure the volume and divide them to
obtain the density something like that. For control variable, I can give you actual example but
then the basic idea for description is you have to explain how you can maintain that variable
to be constant and of course those controls should be meaningful control variable and do not
promise any control that you cannot actively control so for example if you cannot control the
air pressure then I would rather say you don't put it down or you can still include it but put it
at the end of your table so there is you say you always should put the more important control
variable first so then it's just intuitive when people read it they know this is the most
important control variable."

Video 5: Apparatus, Experimental Setup, and Procedures Video Transcript:

"in this video we'll talk about the apparatus, experimental setup, and procedures in your IA.
Apparatus session is quite simple you just have to include all the items that you need but
one thing that usually happens is it takes too much space in your report so if you put down it
as one column like this then you basically waste a lot of space. So I'll recommend you to put
it into two or three column like this one or even this one so that you can start the next
session below here so it will just save you some space. The next thing that I'll recommend is
if you have any instrument for example simply a ruler as I say if you use a ruler try to put the
precision as well so let's say the ROI is plus or minus 0.5 mm all right is there's no harm to
put a bracket and put on the precision of it or maybe an electronic balance and your bracket
plus or minus one gram and that would be helpful and show to the marker that you pay
attention to the precision. Next, experimental setup so according to the Ibo exam report
either a photo or a sketch will be acceptable they even said they prefer to see sketch more
than a photo because a sketch usually it's clearer and it is simple enough cause you should
be doing it with simple lines actually these these will be good and better than this because
our we are not artists and so actually we prefer simple lines either way you should include
the annotation, detail annotation to point out like which part is is what if you go for the photo
approach then you should pay more attention to something like whether your personal
identity like your face is disclosed or it could any mistakes that you accidentally show in a
photo depending on your research question maybe some research question you need to
align the instrument or do anything like align different apparatus maybe you didn't that will
show a mistake or you may be showing a certain that is not safe enough for example maybe
you are doing something with high voltage but then you are doing it next to a what's a tap or
a water source so that is a stupid thing for you to do so make sure when you put down that
photo make sure or actually when you take the photo make sure there is no such thing no
mistake can be observed and the last thing is it just be helpful for the communication that if
you have something that is good elephant in the background so for example if you don't
need a beaker in your experiment so if there's a beacon nearby just put them away and that
will just be much much easier for the marker to see and also understand your context I
guess this is why the whole thing photo is not that prefer while this sketch was actually
accepted in a better way just because the other candidates didn't pay attention to all these
things. Lastly, procedures you want to only put on the key important steps only like I
mentioned in the previous video and shown in the example points well they don't want to see
something like a plug in the cable all right those are not necessary at all right just put on the
key important steps. Next thing is that you want to use a point form with numbering so
maybe that times where you finish a certain step and you go back to change for the next
interval that you can refer to all go repeat step say for to act something like that for the next
uniform they will be much much easier to understand and follow. Next thing is not only above
all procedures but also the design of the Berman's that is you should include at least seven
in the photo independent variable and you should always think about whether you could
repeat each independent variable for three times so that you can get to do the error analysis
later on and you can also justify the error part of the group data later on and last but not least
when you write deep prick procedures some people will just just put down like a coke box
now but actually more what is something we look for is the precaution precaution is talking
about how you can make the data more reliable and so that may involve with the detail when
you collect the data so it could be the page when you collect your data so things like
depending on your research question so sometimes slowly is better sometimes quickly is
better let me give an example maybe when you're doing a certain mechanics research
question maybe you're releasing a certain object through a inclined plank or something
maybe you when you release it slowly because you want the initial speed to be zero then
releasing it slowly will be good in another research question maybe you're doing something
to do with thermo right maybe something to do with the heat or temperature so you may
want to do certain steps more quickly or as quick as possible because by the time maybe
when you transfer something to another place then there must be some heat loss right over
the time there must be some heat loss so then doing it more quickly and reduce that so that
is some sort of precaution you have to consider when you design your experiment and you
should show it in the pre-cana in the procedures as well alignment like I said earlier maybe
you need to tell to the marker that what you have done to align a certain thing if it is
necessary and helpful and lastly if you have to measure something and sometimes there will
be some fluctuation for example very simple thing as in for the electronic balance when you
put a certain thing on it the reading may be changing so it could be like one two point three
four gram and then that last digit this second decimal place may be keep changing is
changing from say or seven to two etc right in that case you have to think about what you
have to do in your data collection i've actually make another video which you can refer to it
what to do with this fluctuation so you can refer to that if you haven't seen it you can find a
link in the upper corner here."

Video 6:

"in this part of video we'll talk about how to write the ethical safety and environmental
concerns. First, ethical in fact I have never encountered any ethical issue in physics probably
you may face more of these in maybe biology or even study psychology so physics shouldn't
be an issue if you know any ethical concern you have you can think of then let me know in
the comment section ok. Safety here's a list of something that you can think about whether it
is applicable to your research questions. The first thing is protective wear so it may be
goggles LabCorp even the gloves as well. Next thing is about the warning signs so if your
experiment is kind of dangerous then you may want to put some sort of warning signs so
that to prevent people from going near your experiment area. The next thing is more directly
what kind of potential danger you may think of regarding your research question so it might
be cracking class maybe alright so maybe things can potentially crack something if you
involve electricity so you may be thinking about our weather it will be close to water source
or not if you're working with something that's really extremely cold or extremely hot then
there should be some measures you should do to prevent yourself getting hurt if you're doing
something to do with maybe mechanics moving very quickly then surely that could be a
potential dangerous one so think about what you can do with it and if you somehow maybe
work with water then you may make the floor or your table very slippery so something you
should talk about in a case where you involve any chemicals I think in physics it shouldn't be
something that is extremely special probably you just do maybe table salt solution sugar
solution zero up or like any chemicals actually no matter what chemical you do you can
always try to find a dative organization to refer to the guideline of that particular chemical so
whether or not is irritating to your skin or whether or not it is flammable etc you should be
able to find the issue quote and cite what they set from a certain organization of course you
should be able to judge whether it's of rotative or not whether it's trustworthy or not in a case
where you may be using something like maybe ladies actually I should say laser if it's too
bright then again there should be a certain label or a certain legal qualification to see
whether it is suitable for you to use or maybe over a certain power rating it will require a
license or require a certain measure to do so you should pay attention to that here as well I
wanted to say what is too loud so maybe it will be hurting your ears or so so maybe there
should be an organization telling you at what decibel of numbness it would hurts your
hearing ability or so so all these things is not just you know mentioned by you but you should
always find the certain organization which has experts giving out guideline onto this alright
so this is something you should do to show you have the attitude to be safe in the
experiment if you really want to show a good performance in this area you should always try
to be more thing of the concern that are more specific to experiment so that is to say for very
general stuff like wearing lab coat it may not be much appreciated right or it may not be that
recipe helpful so in those those reasons probably I would not even bother to put up unless
you got really got nothing to put down and lastly this is something that people usually
overlooked as well do not all phi exaggerate the safety issue for example i have seen in the
past some people when they work within their methodology they may just using one or two
or even free breed decel these are like one sort of D size battery and you know one of the D
cell is actually 1.5 v alright this is a standard and then they try to say oh be careful not to
attach the water with it or else you will get your executor all right with the battery and in fact
a battery cannot kill you all right so if you're saying the battery can it have to kill you then I
don't think it's a good impression because from my perspective you is showing a witness that
you don't understand some basic physics behind so really try to think whether a certain thing
that you mentioned is really dangerous or not then then you should consider whether to
include it in this session. Lastly environment so it should be quite straightforward something
you have to think about is whether or not you can actually reduce the waste from your
design so maybe instead of maybe you're using some sort of chemicals or water can you
actually reuse those water from each trial instead of just getting new stuff for each trial that's
one thing second thing is how do you handle the weight so even if you reuse it at the end
you still have to handle like those chemicals or whatever after the experiment right so how
do you actually handle it after the experiment if you try to dispose it would it be actually safe
to do that so maybe let's say you are doing oil let's say so if you're thinking to pour oil to the
sink would it be actually good for the environment or not or simply good for the hole maybe
school or laboratory structure or not is there any way that you can reuse it in a more
meaningful way in a case where you find there's literally no concerns related to
environmental and even safety as well you could actually put down there is no concern but
you should still put on some more justification for that I will actually say if you try hard there
should be some concern you can talk about throughout your whole research question so
shouldn't be the case where you have literally no concern."

Video 7 :

"in this video we'll talk about how to write the qualitative observation and raw data table. First
of all, for qualitative observation, it is one thing that is required in the marking criteria, and I
find quite often that in the past people always overlook this and did not even include
anything about the qualitative observation. In fact, this was very important, and one thing you
should pay attention to is our usual really think about what kind of quality of oscillation you
could have during the experiment because once you've finished the experiment, you collect
the quantitative data, there's no way you can go back in time to look at what kind of
observation you can make in a qualitative view. So really try to think through the whole
process when you're doing the experiment. The most basic qualitative observation you can
always put down is a brief variation trend of your dependent variable with the independent
variable. So let's say you may find when the independent variable increases, the dependent
variable decrease right or increase, of course, but of course, this is just a qualitative way to
say it, there's no mathematical model you could outline by this point. So this one is
something that you could always put down but I would say this one is the least appreciated
one because this is just too general, too simple and not specific to a particular research
question. Every research question can write that. More importantly, this requires a
higher-order thinking, so the first thing is you should check and observe whether during the
experiment your physics framework will have any contradictions. So that is to say when you
to experiment throughout the whole process, is there any violation of the assumption that
you mix? For example, maybe you are doing mechanics question and for a motion that is
probably going in 1D, so maybe a vertical motion, but they actually go sideways, alright, so
that may affect the result and also your prediction according to the physics framework, or it
could be maybe when you're doing a circuit when you are doing certain trials you find out the
components are having a very severe heating effects, maybe it gets to be very hot so that
could be something that you did not expect in your physics framework also. Or maybe if
you're doing something related to waves, maybe there will be some reflection or refraction
even that you didn't anticipate in your framework as well. So these are the things that you
should think through and that is why when you are doing the framework in the very
beginning of your IA, you also need to know clearly what kind of assumptions that you
should have in order to predict using your physics framework. And if you find those formula
online like from other people's research, you don't just find a formula, you also need to know
how they derive it or what kind of again the assumptions that they took in order to derive that
particular formula. If you can find out anything that actually violated the assumption, that will
be very very useful later on when you try to write the evaluation. The second thing that you
should pay attention to is that when you do the measurement, is there any fluctuation in the
reading? So just like the previous video we talked about the electronic balance, if there's
fluctuation, then you should also mention there is a fluctuation. Also, another important thing
that is throughout the experiment, there should be a certain control variables that you should
take care of, but then are they really constant throughout the whole experiment and that is
the issue, right? Because if the control variable is actually changing even by very little bit,
they may affect your measurement in depending upon your prediction or finding out the
relationship fairly. So monitoring the control variable, for example, you may just simply think
oh in festive temperature of the room or of the air will affect, so maybe you find out at the
beginning the room temperature was 25 degrees Celsius and maybe after that it just
changed to 27 degrees Celsius, that may be something you could also mention as well,
likely become warmer simply. So these are the things that you should also check what kind
of other variables you have in your variable list earlier and then throw the whole process
again, you should check it kept careful and no matter if you find out it is a constant flow on
somehow it change unfortunately, you should also report in this session under the qualitative
observation here.
For the raw data table, we have actually mentioned most of the idea earlier just to reiterate
to you again. So conventionally we'll put for the first column to be the independent variable
and your dependent variable should be on the right-hand side and you should have a few
repeat the trial probably free, usually then you should use the notation as one s2 or s3 in this
case cause it's doing distance right to show this is three different trials. Other than that, even
for your independent variable, you should also include a symbol so later on when you refer
to say here the water level heights you don't have to type water level heights this long worse
every time you just have to refer to H and that would be simple. Of course, you should
remember to put down the units, alright no matter what kind of quantity that you're talking
about, unnecessary this is really dimensionless. The other thing that of course argues you
pay attention and you should include it also is the uncertainty, okay? So for raw data, most of
the time that's the uncertainty will be a constant right because that probably is due to your
measurement instrument like the last or the minimum increment of it, so then that's why it's
usually the same. There is a situation where it may not be the same, in that case, instead of
putting it in the column hat you can actually do a separate column. So for example, if say
lysate distance actually is not having a constant after uncertainty right just let's say then what
you could do is instead of putting this here you can actually just set up one uncertainty called
Delta s 3 let's say if this is for s 3 and then /cm and then you can improve here so say if this
was still 0.2 then yeah that's fine 0.2 maybe for the next one zero point 3 then put it as
silicon free but of course if you do it this way later on you have to explain why it is different
for each different trial. And lastly, like we remind you earlier that for raw data cause they are
of the same precision so along the same column they should have the same decimal point,
alright which we measure in the earlier video again so just a reminder before you move on to
process data. There's one more important thing that you should do and that is to give the
justification of the absolute error and you should do it for each of the after error that you
mentioned earlier. Normally you can follow the rule that you learn in a B physics that is if you
use it and a lot in German for example ruler then you can do the smallest increment divided
by 2 for the absolute error and well for digital then you should know it is just normal the
smallest increment. However, other than following this rule, I would strongly strongly
recommend you to double-check with the instrument manual, of course, if you say oh I'm just
using a ruler and of course there's no manual for your check, however, if you are doing any
kind of Vernie equipment there's always a menu you can refer to and in that case, it will be
very helpful to reinforce the idea to the marker that you you want them to give you a better
personal engagement mark because if you are really into this topic then you want to do
everything carefully and check everything carefully. So let me show you an example of how
you can check the manual for the vernier equipment and it could be the same if your
equipment is not on phone yeah there should be a certain menu go along with the
equipment itself. Okay so on Google you can actually type the sensor that you use
expression for vernier so I'll type say temperature probe vernier and then you should be able
to find out different pages so usually the first few will be relevant one so I'll I'll go for the first
one and then once you finish loading you see a double check whether the picture is really
the one that you use of course and after that you can scroll down and there should be some
tabs you can click say specification should be the one that is relevant here then you can see
that it will tell you actually four different range of temperature they have different resolution
that means the smallest increment at the same time they will also tell you another thing that
is called accuracy and that is to say at a certain temperature may be the random uncertainty
should actually be treated as plus or minus 0.2 or plus or minus 0.5 degrees Celsius instead
so that may be much larger than you think and I would say it's really up to you whether you
want to follow this one or the one that you already know thing off however mentioning the
fact that you have try and check this will be very helpful again for your personal engagement
and more scientifically speaking if the menu here said oh actually accuracy is just 0.2
degrees Celsius then I guess you cannot take it up to say 0.03 instead because you should
always take the larger uncertainty when you're doing a research however in some cases I
know that if you take this one then your error bar will be extremely large and you can know
did use anything then in that case you may need to do a small discussion on this why you
want to stick with these or why you want to stick with the resolution itself and that will be
appreciated. And lastly like I said in the previous slide if there's any fluctuation in the reading
so for example when you do the electronic balance is one two point three four but then
sometimes it changed to seven like three seven sometimes it's change to three - lasse so
what you can do is you can either in your methodology says you will observe and take the
maximum and minimum value for a short period so maybe like five to ten seconds and then
say in this case you may not even get level six or seven then say in this case that will be in
this case will be twelve point three seven and twelve point three two right will be a maximum
minimum so that you can use the half range method to find the answer uncertainty in this
case instead of just saying oh 0.01 is the Efrain certainty in fact you can actually add them
up together because when you read 3 7 or 3 - it could be plus or minus 0.01 so then you
should add the range of that the other way of doing it is if this is like the case then if you are
happy with you can just take what whatever you have earlier and it gets the last digit cost
since it is unstable you can just ignore it and then you can take twelve point three instead so
both ways would be fine it's really up to you how you would like to do it

Video 8 :

"in this session we'll talk about how to write the process data and uncertainties session here
is an example of the process data table for my previous student as you can see the
formatting is very similar to the raw data table where you will improve of course the quantity
and also this symbol with the units for all the things that you want to talk about and there are
times where you may have the answer and certainty to be the same for the same quantity
then you can just put it in a column pet like raw data table in a case where you have different
percentage error or absolute error then you can put it into a separate column which of
course this value will be calculated one byte in a case maybe your dependent variable
cannot be measured at rest late for example if I want to find the density as my dependent
variable I probably have to measure the mass and also measure the volume first and
therefore I can calculate each trials density and then I can put it into the processed data
other than that you may also need to think about later on when you try to plot a graph you
want to plot a linearized graph and so what you do is when you look at your framework you
should find out the relationship between your independent variable and dependent variable
how they can be linearized so for this student he finds out earlier that somehow tell sorry
square roots H is proportional to s so that's why you can see in his process data table he
deliberately want to find the square root of H even though he measure H only at the
beginning in the raw data table if you don't know how to linearize the equation that you find
try to watch the video from Chris Doner I'll put a link in the description below after you finish
the processed data table you should also provide the or calculation especially for error
propagation so showing how you can calculate all the values in your table for some simple
calculation like all simply square rooting the edge those would not be required even in the IV
exam they said taking the air fridge is not required although I would say if you have space
then you can still in crew it most of Uncertainties the propagation you can actually use the
equations that you learn in chapter one to deal with it so here I will not explain this again
since you have learned in Chapter one in some cases however you may be having some
other special mathematical function for example trigonometry like sine function or lot function
or exponential function then these uncertainties rule may not apply that rest that you cannot
use actually so let me teach you how you can do that to find the uncertainties of those
function if you try to find out the equation that is to find the area of these kind of special
function then you should find something like this a chart like this so you can quote the source
of this and you can simply use the error equation alright so you may not understand why it is
the equation like this because that involves something in mathematics quality partial
differentiation in fact the equations you learned in Chapter one also required partial
differentiation is actually approximation so you did not actually encourage and account for
100% of the error it's just a good enough deal of absolute error only anyway so let's take a
look here it means that if you have loss function or its potential function or discount to
Kalama tree function you will be sticking with this alright so f is like the function itself and
then a is whatever still into that function so let's say if you do sign function then FX will equal
to sine X and let's say you have a certain measurement that is 30 degree plus or minus 0.5
degree using the popper method let me just have a proper method first what you do
according to our method here's Delta F that means the absentee of the whole thing psy X
would equal to Delta a times cosine a so in that case it would be cosine 30-degree times 0.5
degree in Radian and that will be equal to this value so 7.5 something times 10 to power of
negative 3 this is what I call the popper method okay so I would say if you do do it this way
as a marker I'll actually appreciate you more because you show me how like you personally
engage and try to find a more accurate way to define the error if you're not mathematically
strong or you don't want to spend too much time to to deal with this there's another
alternative I would say it's also being acceptable in IP physics and I call this an intuitive
method all right so the intuitive method is really intuitive so what you do is think about that
sine X exists what you want to find at the end and when you do the measurement X is plus
or minus 0.5 degrees so that is you say X could be in the worst case thirty point five degree
or twenty nine point five degree and so what you can do since you know sine function is a
monotonic function within it is at least within 90 degrees so you can simply substitute these
two value in the way you'll find the maximum and minimum bound of sine X for this value so
that will be in the range of these two and so what you can do is you can simply use the
so-called half wrench method that we learn in chapter 1 which is just minus them divided by
2 and actually you'll find you can find a value of these which to be honest is very close
enough so I would say both way I would say both weight will be accepted and if you could do
the proper method then to the common method if not then you can stick with the intuitive so
that will be same for the potential and log function as well."

Video 9:
in this part of video we'll talk about how to do the graph interpretation and draw the
conclusion from it. It is one of the hardest, toughest part of your physics IA. The easiest part
is actually plotting the graph, you just have to make sure that you must have included the
error bar for both X and Y axes. There are times where maybe the error bar is just too small
to be seen because of the absolute uncertainty is relatively too small, so you have to
mention this specifically. Don’t just skip it, you have to really mention this in your passage.
And also, if you could show the numbers you explained, so maybe saying that the range that
you are doing of the data maybe it's like 1,000 but then the uncertainty of Y is around maybe
0.1, so there's no way that you can show it in the graph in a visible way.
Next, when you turn graph, there could be mainly two types of graphs. One is linearized
graph, so that is to say you successfully find out the framework equation and you find a way
to linearize the equation itself. So like the previous video I said in the process data table, you
should have prepared those values and so you can just plot it out. Alright, so this should be
the most straightforward way. For example, if you already find out the Y is proportional to 1
over X, then you can simply of course present the data or the values of the data in the
processed data table and then you plot Y against 1 over X, and probably you get a straight
line, hopefully. So this will be the most straightforward way to do, and in that case, you
should also be able to find out the max and min line as well. Okay, max and min line is only
available when you linearize the graph or the relationship in itself is linear.
Other than that, when you draw the max and min line, you should make sure the lines will
pass through all the data points with this error part. Some people may not understand
capacity and think that when I draw the max and min line, I just have to locate the two points
or I will jot this point and this point is for a line and this point and this point is draw a line.
This is a simplified way to do it, but then when you draw this you should actually pay
attention to whether those error bars in between them, whether or not it pass through them
as well. So if it doesn't, then maybe you should adjust the point a bit higher or lower to make
sure that the line max and min lines do pass through all the error parts of the other data
points to validate the max and min line itself. So don’t be simple-minded and just simply pick
the four coordinate and plot it out.
In case if you really cannot linearize the equation and plot a graph, to be honest, this is still
fine to plot the independent variable against the dependent variable. It may just be hard later
on when you're trying to analyze the validity of your graph later on. If you don’t know how to
linearize it, you can try your luck by maybe finding out the value and plotting log of your
independent variable against log function of the dependent variable. So if the true
relationship of the things that you're studying is in power equation, that is Y equals to A times
X to the power B, then you should be able to find a linear line. However, this may not be
always the case, right? Then it won’t work.
In case you have different error bars, that means different uncertainties for each data point,
you can refer to this video and learn how to plot the graph on the Google sheet. Of course,
you can use other software like Excel or even Vernier Logger Pro. This is just a link that I put
in the description, you can find the other software on YouTube easily as well.
Interpretation of the graph is the hardest part. Here an idea related to TOK is that science is
actually not completely objective. Maybe in primary or secondary junior secondary you will
feel that all science is really good because it's very objective. However, in a case where
maybe when you're interpreting the graph or interpreting the data, you could have different
ways. Just like us when you look at random painting, you can have different feelings. Of
course, it's not directly relying on feeling. However, there could be more than one
explanation and it really depends on whether you have justification for your argument. So,
there is no standard method here.
Before you draw your line of best fit, you may find out that there might be some anomalies
and it is actually fine to have that in your physics IA. However, if you have that and if you
want to declare there is anomaly, then you should have a proper explanation for it. As for the
quantity, I would say if you have one out of seven dataset it is acceptable and if you have
two of ten dataset it is okay. But if you have more than that, I guess you may want to revisit
your methodology or revisit at least that particular data point and then hopefully maybe you
can try to redo and repeat more trials on that point and see whether you can find something
different.
After you get all the error bars and data point ready, then you can decide what kind of line of
best fit you can put to represent the trend. There could be more than one possibility. For
example, in this case, you can actually fit a straight line, of course, at the same time you
could also fit a curve, actually right? Which maybe cut out the equation or it could be
exponential function, etc. No matter how many you can do, at the end you should only
choose one of them. And in order to choose one of them, you have to make an argument to
see, oh how come I will choose the for example the exponential one, the blue one here,
rather than the straight line. So maybe that's something to do with the framework that you
derive earlier or it could be related to the literature result. Maybe you say, oh this match with
other people's research while the linear was not actually found in anyone's research. The
other way of making an argument could be try to extrapolate the trend. So and then you can
see whether there’s any unreasonable result. For example, let's say in this research question
I'm trying to find refractive index, let's just say. There may be if you're using a linear
relationship then it may not make sense because maybe when you try to draw a line like this
that like keep extending it or if you keep extending that the way that the index can go up to
like five or six and that doesn't make sense. I mean no matter what material you do, probably
it shouldn’t exceed two point something or three for something. So this is a way to extract
your extrapolated extreme point I can help you to identify whether a certain mathematical
model will be suitable.
So after the arguments that you made earlier, then in the conclusion session you should be
able to state clearly what kind of mathematical model will be the best for describing the
relationship of IV and DV that you're studying. So do not stop at just saying oh when this
increases the other one decreasing right, this one would not get you anywhere or as popular
just level two or three if you just say this way. You really have to describe it by using the
terminology in maps, that is probably linear, proportional, exponential, inverse, inverse
square etc. There's something that I would like to remind you that is, do not get confused
with linear and proportional. Actually, I mentioned in other video as well but let me just
reiterate once again quickly, linear the equation is y equal to K X plus C. Now for
proportional it is y equal to K X, so proportional actually pass through the origin. Well for
inverse function, sometimes they get confused with being negative linear. So for inverse, it
will be y equal to K over X. Well for negative linear could be y equals to negative K X plus C,
of course K will be a positive in number here. In terms of the shape, both actually for inverse
and negative linear, they will have relationship like when Y increase X will decrease, but then
in terms of the shape of the graph, inverse is more like a curve like this while for negative
linear line might be something like this right, it's a straight line. So that's why you see like
what if you just say or when I increase X decrease that would not be enough, you really
need to use a clear mathematical model and there is a function to show the relationship
between the independent variable and dependent variable.
And here is something more you can include in the conclusion, some people may like to
include in the evaluation instead. I would say if you include the conclusion that's actually fine
as well. If you have a linearized linearized graph, you could actually discuss the physical
meaning of the slope because think about, you usually have the mass mean line right? So
mass mean my probably something like this, and then there should be two slopes that you'll
find, you should be able to find uncertainty by using the half range method. If the slope has a
physical meaning that that means this uncertainty has a physical meaning too. So maybe
representing a certain physical quantities uncertainty. So try to find out, you may find useful
later a lot when you do the evaluation for comparing with the literature. Same applies to the
intercept, so for either X or Y intercept, their uncertainty may be meaningful. So let's say if I
stick to this graph, then obviously X NSF will be here and here. So you want to find this as
the uncertainty, or if you extend a line like this, what this is a terrible one right, extend the
line like this then you could find this is a whiners have, this awareness at. So this is the
uncertainty, you could divide by two for the Y intercept and if again if these things were
percent a certain meaning name physics then surely you should point it out and maybe
discuss here or later the other thing that you can do with the linearized graph is you can also
check this origin and that you should say if you expect the line of best fit to be proportional or
if you can simply just look at the physics and see when this is zero whether or not the other
one will also be zero if it is not logically true then you may want to check whether your data
will cover the origin so that is you check whether the mass mean line so for example if the
line something like this and something like this then it is fine because this is the area where
you cover so the origin is within the area or if you think about it if these get too then it doesn't
cover the origin then in this case it will show maybe there is somatic era you should think
about later. Common mistakes in conclusion on to more common mistake that I would like to
remind you that I find quite a lot in the conclusion section that is some people would say oh I
prove something so no matter how nice the train that you'll find how well it match with the
literature you should not say proof and that's something related to tea okay as well in
science you never prove anything because what you do an experiment is to provide
evidence to support a certain theory or a principle or a scientific model so you can say were
site you verify something you let a date something or you have something that strongly
suggests the model is correct those will be the words that you could use it's always easy to
disprove something so that's why proof is is like a work that we we don't want even want to
touch we don't have the confidence to say we prove something cause you never know
maybe a thousand years later like Newton's law that you'll find out all actually there's some
flaw so try to avoid using this work the next thing is some people I don't know why they
would just like claim the random error is rather large or small by your own feeling so you kind
of do it in a way that all when I see the error bias baked that I think it is big so there you just
say lost so this is not actually a way to do it in a scientific way this scientific way to do it is
whether or not you can fit different mathematical models or just like what I showed you
earlier if you have the error bar that is so huge that you can fit a video line and then maybe
we can fit a curve to a different mathematical model then quite likely your random error is
large because think about this if your error if your random error is small that is maybe it is so
so so so tiny then the only way to do is probably really a straight line right there's no way you
can't fit a curve I don't think you can you can think of a curve to fit I mean yeah really there's
no way so that's why we always want to reduce the error by itself if possible and that could
be rich of course other than the enjoyment itself maybe your methodology on how to set the
control variable or how to collect the data it will be important as well so remember whenever
you claim whether the random error is large or not you have to base on the evidence."

Video 10:

let's move on to comparison with the

0:03
literature
0:04
the first question that you may ask me
0:06
is should you include the literature
0:08
in the introduction or framework session
0:10
or in this section
0:13
because if you pay attention to the
0:14
structure that i introduced to you you
0:16
will find
0:17
the introduction and framework should be
0:19
done in like the second part
0:20
of your ia so should you do it here
0:24
or should you do it at the end
4 possible “literature”
0:27
the answer is it depends so it depends
0:29
on what kind of
0:30
literature that you are using and there
0:32
are four main
0:34
possible literature that you may use for
0:36
your physics ia
0:38
the first one is called the formula
0:40
which you may find just simply from the
0:42
source somewhere
0:43
or you can derive it from the
0:45
fundamental law like f equal to m a and
0:47
you may be combined with several
0:49
equations and you find out
0:50
an equation at the end so this one you
0:54
should put it into the framework section
0:56
or you can also include the microscopic
1:00
view
1:00
which is maybe the particle view so if
1:02
it is related to wave
1:04
uh it can be about wave as well so
1:07
that could be included into the
1:10
background or framework also so that
1:13
means the
1:14
very early section of your ia because
1:16
without this
1:17
basically it is very hard to understand
1:20
your methodology and all the things that
1:21
you did earlier
1:23
however if you want to use other
1:25
researchers experiments or results
1:28
or even in a more last
1:32
resort or desperate situation
1:35
i would say you can also use simulation
1:37
for comparison
1:38
so for these two number three and four
1:41
you could
1:42
leave it for the end so in only in you
1:45
in this section when you compare the
1:47
literature then you can mention it
1:49
so in this video we'll talk about what
1:51
to do with each of these
1:53
literature one by one okay so the first
Formula / derived framework
1:56
one is
1:57
called the formula or derived framework
2:00
to illustrate you a better idea i'll use
2:02
an example
2:03
so let's say the formula that we use is
2:05
the gravitational law
2:07
f equals to g m one m two over
2:10
r squared so uh i chose this because
2:13
probably you won't be doing this
2:15
i mean not many people would do this
2:16
because this is not quite doable
2:19
as you know uh the gravitational force
2:21
between two
2:22
normal objects will be very very small
2:24
that you can't really measure with the
2:26
tool so i guess this is a good
2:28
way for me to not to give you an actual
2:30
formula that you might be using
2:33
but use it as an example the way that
2:36
you do as long as you have your own
2:37
formula
2:38
should be very very similar as long as
2:40
you follow my steps here
2:42
so the first thing that you do
2:45
let's say you already finished your data
2:47
collection and start doing this section
2:49
and
2:50
what you have to do in this approach is
2:53
you have to measure the control variable
2:55
so let's say
2:56
in my this i report uh or study
3:00
i'm i'm doing the independent variable
3:02
as
3:03
the radius so that means like a distance
3:06
between two objects
3:07
and the dependent variable is the
3:09
gravitational force
3:11
i just let's say i just magically
3:14
measure it
3:15
okay so um what i have to do
3:18
here is i have to firstly measure the
3:20
control variable so obviously
3:22
m1 and m2 the two object that i use will
3:25
be control variable
3:27
as well as the constant itself is
3:30
constant of course
3:31
so what i can do is for these two masks
3:34
i could use a maybe electronic balance
3:36
to measure it so maybe i find out oh is
3:38
uh
3:39
literally hun one kg or the other one is
3:43
1.002 kg let's say
3:46
okay so in this case then i would have a
3:49
number
3:50
and of of course the uncertainty also so
3:52
maybe that would be
3:55
0.0001 kg and this is also the same
3:59
all right etc so for
4:02
some values that you cannot measure
4:05
like depend again depends on what kind
4:07
of ia topics that you are doing
4:10
some of them just can't be measured
4:11
directly or it depends on the material
4:14
itself so like this one g then you can
4:16
simply
4:17
research to find of course trustworthy
4:20
source
4:22
on the value of this so you can do that
4:24
as well
4:25
and of course if you could try to find
4:28
out the
4:30
error uncertainty absolute uncertainty
4:31
of it also so you can
4:33
use it later on if you really cannot
4:35
find then maybe you can use the last
4:37
digit
4:39
but that is not very ideal of course
4:42
so if you can find out somewhere you can
4:45
mention
4:45
the abstract journey that would be great
4:48
step number two
4:49
what you have to do is then you can
4:51
calculate
4:53
using the formula you can find you can
4:56
calculate the corresponding
4:59
dependent variable with the independent
5:01
variable so independent variable
5:02
in your original research so let's say
5:05
for our i chose
5:07
to be say one meter 2
5:10
3 to all the way to 4 to 7 meter or 10
5:14
meter let's say
5:16
okay so this is something i chose so
5:19
it's
5:19
perfectly fine you can still use this
5:22
independent variable to substitute
5:24
into the formula so you can substitute
5:26
into the l square here
5:28
and you should be able to get the
5:30
dependent variable which is the force
5:32
here because you already got other
5:35
values in your formula which are the
5:37
control variables of course
5:39
so in this case you'll be able to
5:40
generate a table
5:43
of your independent variable and
5:44
dependent variable like what you do for
5:46
your data processing in the earlier
5:48
section
5:49
so that's step two step number three is
5:52
again similar to what you did for your
5:55
own
5:55
data you can also do the absolute
5:57
uncertainty
5:59
uh you can do calculation uh that means
6:02
doing the
6:02
error propagation method by
6:07
all the absolute errors you know so like
6:09
what you learn in charter one or
6:10
basically what you did in the earlier
6:12
session for your own
6:13
experiment data you will be able to find
6:16
the error
6:17
of force that means a dependent variable
6:20
here
6:21
um the reason is because at the end even
6:23
though this is from the literature
6:25
uh it would be nice if you can get the
6:28
error bars also
6:29
it's not really about the formula itself
6:31
is more about
6:32
maybe just maybe let's say the mass that
6:36
i use i find out maybe
6:38
uh the way that i measure is terrible so
6:41
let's say
6:42
let's say it is not uh like what i said
6:45
let's say it is something else like um
6:48
say
6:50
it is one point two plus or minus zero
6:53
point
6:54
five no let's say 0.1 okay
6:57
and the other one is 0.2 plus or minus
7:01
0.1 let's say
7:02
okay i know you won't do it but let's
7:04
say then
7:06
even though with this and the so-called
7:08
literature
7:09
equation you will still generate a large
7:12
error but for sure especially you can
7:13
see
7:14
this one all right so we can still
7:17
consider the impact
7:18
of the error uncertainties
7:22
and lastly of course the ultimate goal
7:25
of obtaining the
7:27
abstraction is to plot it onto the graph
7:29
of course so plot
7:30
the data points so from the data table
7:33
again using
7:34
the literature uh the formula to
7:36
calculate
7:38
the theoretical value shall we say
7:41
so that we can plot the data point
7:42
together with
7:44
its error bar so maybe in the data table
7:46
you can also add
7:48
the uncertainty of the dependent
7:49
variable also
7:51
as well as for independent variable you
7:52
can also say it
7:55
because that's also measured from your
7:57
experiment too
8:00
and you can plot all these things onto
8:02
the same graph
8:03
with your experimental data so the
8:07
main idea is that when you do this
8:10
then you can compare your experimental
8:13
data with the theoretical data because
8:16
the experimental data for example back
8:18
to my
8:18
equation here is that i measure the out
8:21
and measure the
8:22
f right because i mean this equation
8:25
might be wrong or maybe there there are
8:27
something that
8:28
affect my measurement right so i measure
8:31
these for experimental but then when i
8:33
do different radical i
8:35
calculate from the equation so that is
8:38
a different approach for sure so we'll
8:40
try to see whether these two approach
8:42
would
8:42
show you the same relationship okay so
8:45
let me just
8:45
illustrate you the graph and give you an
8:47
example on
8:49
how it might look like so back to our
8:52
formula here
8:54
if we want to generate a linearized
8:57
graph
8:58
quite likely you have f to be the y-axis
9:01
and x-axis should be 1 over r squared
9:05
and let's say in that case that you
9:07
probably should get a
9:08
straight line in your experiment so
9:11
let's say
9:11
it is something like this and look at
9:15
the slope well this is experimental data
9:18
let me just write down here
9:20
and note that the slope actually means
9:23
this one right g
9:24
times m1 m2 that is the physical meaning
9:28
of the slope and well let's just say
9:32
according to your error bar you can get
9:33
the mass mean line to be
9:35
something like these let's say okay so
9:39
you have an area of these and let's just
9:43
say the theoretical one
9:44
is uh somehow like
9:48
this okay let's say all right the
9:50
theoretical one
9:51
the blue one okay so uh this is where
9:55
the discussion starts
9:56
and so for the first thing that you
9:58
should consider check
9:59
and discuss is whether or not the
10:02
theoretical that means the literature
10:04
line of best fit the blue line will line
10:06
within
10:07
your experimental mass mean line range
10:09
that means the red
10:10
range here so the answer is obviously no
10:14
in my example because you can see yes
10:16
there are part of it
10:18
they overlap each other but then you can
10:20
see here it doesn't overlap at
10:21
all so there must be some
10:25
problem or some disagreement between the
10:27
theoretical and experimental
10:29
data so it's really up to you to judge
10:32
whether or not there is an offset of x
10:35
or y
10:36
that means translation in mathematics or
10:39
offset of
10:40
the slope in itself um you should
10:43
not just think about the mass but also
10:45
think about whether it
10:47
fit to the physical idea in your
10:50
experiment
10:51
so for example back to our example here
10:54
you may judge that just may okay
10:58
not necessarily you may say uh
11:01
the experimental data if it becomes
11:04
steeper
11:05
like this then it will more match with
11:08
the theoretical data and maybe translate
11:11
to the left side
11:12
so then that may be something to do with
11:15
its slope
11:16
first of all so that is to say uh you
11:18
have to look into its physical meaning
11:20
as in
11:21
as we said slope represent g times m1
11:24
times
11:25
m2 so maybe there's overestimation of
11:28
underestimation of m1 and m2
11:30
when you try to generate theoretical
11:32
value and you have to see
11:34
uh since vertical value is more steep
11:36
right so then the slope is
11:38
overestimated compared to the
11:40
experimental one so maybe
11:42
you overestimated m1 or m2
11:46
in this case so there are different ways
11:48
of working it out and
11:49
you really have to judge case by case
11:52
and
11:53
it really depends on how your data look
11:55
like but these are some basic ideas
Microscopic view
11:58
next the second thing that you can do is
12:00
the microwave view so that is to check
12:02
the particle theory
12:04
so in your background research
12:07
like the very early part of your ia you
12:10
should have done this already
12:11
and here is you just double check again
12:14
whether or not
12:14
uh the theory will match with your
12:17
result qualitatively
12:19
so that is to say maybe when your
12:21
independent variable
12:22
increase how the dependent variable
12:25
should react
12:26
in terms of values so in some cases when
12:30
you have discrepancy
12:31
in your graph it may also be very
12:33
helpful
12:35
to explain the differences so for
12:37
example
12:38
i had a student uh and the graph
12:41
was something like this
12:44
okay and the expected line is more like
12:48
these all right so the uh the
12:51
theoretical watch label here all right
12:53
the theoretical and the red one is the
12:56
experimental
12:57
so we can see that the main discrepancy
13:01
is around here and here so the middle
13:03
part is all good
13:05
right you can see nothing's wrong like i
13:07
mean of course i mean if i could draw
13:08
the
13:09
uh max min liner it would be even better
13:12
so the
13:12
main thing is uh something is wrong at
13:15
the beginning or at the end
13:16
so maybe you can think about the
13:18
particle view
13:20
something maybe certain side effects
13:24
in terms of the particles may happen and
13:26
that
13:27
may affect your result
13:30
or maybe there are some assumptions that
13:32
got violated
13:33
or become more significant in terms of
13:36
the theory itself literature number
Other researcher’s experiment results
13:40
three
13:40
so this is something that you could
13:42
start to mention here
13:44
and that is someone else's research the
13:46
first thing that you should pay
13:48
attention is that
13:49
of course the source of these experiment
13:51
results should be trustworthy
13:53
so it cannot be something like another
13:55
ib students
13:56
ia or ed report ideally at least
14:00
it should be from the university so it
14:02
cannot be any kind of like blog
14:04
or wikipedia or any like a random
14:08
website at all
14:10
there should be the name of the
14:12
researchers
14:13
and also the universities name also
14:17
the second thing that you should pay
14:18
attention is that the range of the
14:20
independent variable uh that should
14:22
cover the one that you have
14:24
of course uh usually they do wider than
14:26
yours because
14:27
they are just more resourceful and also
14:30
pay attention to their assumptions so
14:32
maybe there are things that
14:33
they have assumed and you don't or the
14:35
other way around and that is
14:37
also worth discussion as well the third
14:40
thing that you you may want to
14:41
pay attention to is the control variable
14:44
so maybe
14:45
you are using a different kind of
14:48
size that may also bring an effect or
14:51
even the methodology itself is different
14:54
so that's something that uh i mean this
14:57
i'm not the thing that uh saying oh if
14:59
you are having a
15:00
different uh assumption or different
15:03
control variable etc
15:04
then you cannot use it or it's more
15:07
about
15:07
uh of course the research question
15:09
should be similar or close and relevant
15:11
together
15:12
but then maybe there are some of these
15:14
that makes a difference
15:16
uh how do you justify how you compare
15:21
those differences and rule out maybe
15:24
that would make a difference also
15:26
more importantly after like you check
15:28
all these things
15:29
maybe you find out you can finally use
15:31
it then you can check whether or not
15:34
the mathematical relationship so once
15:36
again not just
15:37
increase or decrease it's more about the
15:39
exact mathematical
15:41
model so linear exponential quadratic
15:45
inverse square etc of the independent
15:47
variable and dependent variable
15:50
whether or not is matching yours and
15:53
their results and you can also do this
15:55
similar thing
15:56
like what we did earlier instead of
15:58
using the
16:00
formula that we have to plot the
16:03
theoretical value you can also use
16:05
their theoretical value to generate the
16:08
same graph
16:10
and you compare like what we just
16:12
mentioned earlier
16:13
so that in that case uh you can do the
16:15
same thing and you try to find out
16:17
maybe there's some discrepancy and you
16:18
also discuss why there is a discrepancy
16:23
there are some cases where you may not
16:25
be able to
16:26
find a clear mathematical relationship
16:29
mentioned
16:30
by the other researcher or even or even
16:33
the slope itself so what you can do is
16:35
at least probably you can at least find
16:37
the dependent variable
16:38
and so you what you can do is you can
16:40
plot it on to again the same graph and
16:42
so you can then
16:43
see maybe the line of yours is
16:46
more like the blue one and the red one
16:49
is the
16:51
other researchers experimental data so
16:53
they can kind of compare them
16:55
and also if you think there's a clear
16:57
trend and of course you can also draw a
16:59
line
17:00
for their data also or at least you can
17:02
try to compare
17:04
yours with all the data that they have
Simulation
17:09
okay so we have covered the three
17:12
main kinds of portable literatures
17:15
and i would strongly recommend you to
17:17
cover all these three
17:19
in your ia so for most cases you should
17:21
be able to do it
17:23
there are some cases where uh say there
17:25
were some students in the past
17:27
told me they could not find other
17:29
researchers experiment
17:31
of course a trustworthy source then i
17:34
would say you
17:35
may also consider using simulation as an
17:38
alternative
17:39
for number three so uh in case you're
17:43
using this
17:44
then you should note that um you cannot
17:47
generate error bar
17:48
in the simulation so that may cause an
17:50
issue but then the rest should be
17:52
uh similar to what we mentioned so you
17:54
still generate
17:55
uh a trend and you plot them onto the
17:58
same graph like what we did
18:00
earlier and you can still do the
18:01
comparison but once again if you could
18:04
just do the first three it shouldn't be
18:07
that hard to find a relevant researchers
18:10
experiment
18:12
okay so that's pretty much it and lastly
18:15
once again i want to reiterate
18:16
you should always always consider
18:20
the physical reason okay i know this
18:22
part involved quite a lot of maths
18:24
but ultimately it's not a math is a
18:27
physics ia so you always
18:29
should link and explain why there's a
18:31
discrepancy
18:33
using the physical concept or
18:36
ideas

Video 11:

finally we are at the last part of our

0:03
ia report and that is called the source
0:05
of error and limitation
0:06
and also the improvement and extension
0:10
first of all i would like to talk about
0:12
the format and here is an example from
0:14
the ia
0:15
sample from the teacher supporting
0:17
material i'll put a link in the
0:18
description below
0:20
and you will find the file there if you
0:22
go to the link you see the same file as
0:24
i'm showing you
0:25
here and you can read the details if you
0:28
want but what i want to show you
0:30
is at the end in the evaluation i would
0:33
like you
0:34
to have the same format as well so
0:37
mainly i would say it is more about
0:40
showing it in a table
0:41
rather than in paragraphs because it's
0:43
just much easier for you to follow
0:45
and also for the examiner to follow as
0:48
well
0:49
so note that there will be free column
0:51
one is the source of error
0:53
and it's effect and also whether or not
0:56
it is
0:56
significant and how to improve it
Methodological vs Procedural
1:02
before we start to talk about what kind
1:04
of errors you can put on you have to
1:05
understand the difference between
1:07
procedural and methodological
1:09
error first which is written in the
1:11
marking criteria and is also mentioned
1:14
in the exam report
1:15
which i mentioned earlier in the
1:16
previous video so you can see
1:18
the word procedure here and also
1:22
methodological issue
1:24
in the middle and the higher band so if
1:26
you really want to get a high marks
1:29
in evaluation you should focus more on
1:31
the methodological
1:33
issues so you may ask what exactly are
1:36
they
1:37
and here is a brief idea for procedure
1:40
which is the one that
1:41
only can give you the lower band
1:44
is involving ideas like repeating
1:48
more times or use a more precise
1:50
equipment
1:51
so you can see that these are examples
1:53
of the so called error that you can
1:55
actually apply to
1:56
all the research question so basically
2:00
everyone can use it
2:01
and that's why it's not being
2:03
appreciated as much
2:05
as the methodological because for the
2:08
other one the methodological one is very
2:12
research question specific okay so it
2:15
depends on what kind of
2:16
ia what kind of investigation that you
2:19
are doing
2:20
and it is also involving the design of
2:23
the methodology
2:24
so that's why i can't tell you directly
2:26
what kind of error you have you have
2:27
really have to think about
2:29
what kind of error you can realize or
2:32
discover
2:34
but of course i'll give you more hints
2:36
in the coming few minutes
Errors and limitations
2:40
first of all for error and limitation
2:42
you should put the
2:43
most significant error or limitation
2:46
first
2:46
in your table because there might be
2:49
some other errors
2:50
but they are not as important so i would
2:52
say uh you should always put the one
2:54
that has the
2:56
biggest effect to your eye report first
2:59
the next thing is uh you may want to
3:01
start brainstorming the error
3:03
by checking the assumptions in your
3:06
framework so whether or not
3:08
the assumption are still valid
3:10
throughout the whole experiment
3:11
maybe it is violated when you assume for
3:14
example when there is no
3:16
air resistance but of course there might
3:18
might be
3:20
a certain effect so these are the things
3:22
that you could check
3:23
by going back to look at the framework
3:26
assumption that you made
3:28
the second thing you can check is
3:29
whether or not your dependent variable
3:32
is really depends on your independent
3:35
variable
3:35
only because there might be a case where
3:39
you your independent variable change
3:41
something else and that thing
3:43
that variable parameter will change the
3:46
independent variable
3:47
indirectly so that is something you may
3:49
want to
3:50
check and read through as well this one
3:52
is quite tricky you may want to check
3:54
in some other literature carefully and
3:57
lastly something you can check is
3:59
whether or not your control variable is
4:02
really constant of course i mean ideally
4:04
it should be constant
4:06
but whether or not it is really constant
4:07
or to what degree
4:09
it is being inconsistent that is
4:11
something you could also mention
4:13
as well as for the two kinds of error of
4:16
course as you learn in ibm physics you
4:18
know there will be random error and
4:19
systematic error and for random error
4:22
usually people will talk about their
4:24
equipment because this is where you have
4:25
your random error
4:27
due to the uncertainty of the equipment
4:30
however we don't appreciate if you just
4:33
simply blame
4:34
the equipment for the precision because
4:36
for every equipment
4:37
they have precision issue even for
4:40
equipment in nasa or in whatever science
4:43
lab
4:44
the equipment they use they must have
4:47
a certain absolute uncertainty so it is
4:50
not about
4:51
realizing hey they have there is
4:53
absolute uncertainty it's more about
4:55
whether that absolute uncertainty is
4:57
large or not so
4:59
you could talk about it but don't always
5:01
talk about it if it is not significant
5:03
so you can try to see maybe depending on
5:07
your measurement if your measurement
5:09
somehow
5:10
uh is very close in terms of percentage
5:12
like i
5:14
took the example earlier if you're
5:15
measuring 0.2
5:17
plus or minus 0.1 of something then
5:20
of course that's going to be a large
5:22
random era or
5:24
if you have having like zero point uh
5:26
four like this is about fifty percent
5:28
right i would say
5:29
uh for around maybe fifteen
5:32
or ten percent is still quite large but
5:34
if you are within say ten percent
5:37
um i guess it's not something you could
5:40
you should talk about uh for the first
5:43
priority
5:45
and lastly for the systematic error uh
5:48
since you know systematic error in this
5:50
nature it should have the same
5:52
direction of the effects so that is to
5:55
say when you say measure something
5:57
for example one of the example that you
6:00
should have learned is
6:01
it could be when you measure a sphere uh
6:06
you may not be able to measure
6:08
the diameter that is the largest length
6:11
basically
6:12
as possible you may be measuring
6:14
something else like
6:16
this one all right so within this only
6:19
so
6:20
you are underestimating the diameter so
6:23
you may want to talk about how
6:24
the underestimation of this would lead
6:26
to something else so maybe you
6:28
underestimate the
6:31
speed or whatsoever so my point to make
6:35
is that you have to
6:36
specify how this direction
6:39
of the effects are measuring more or
6:42
less will affecting
6:43
other measurement and eventually lead to
6:45
your dependent variables whether or not
6:48
the dependent variable will be
6:49
overestimated or underestimated
6:52
this should be specified instead of just
6:54
saying oh
6:55
it will be affected all right so that
6:57
will be very vague and we don't like it
Possible improvement / future extension
7:00
for the possible improvement and future
7:02
extension these are something that
7:04
you could think about however it does
7:07
not cover everything so
7:08
still try to train more after going
7:12
through this list
7:13
so first of all uh i would say you
7:15
should not
7:16
try to suggest something when you when
7:19
you're making an improvement
7:20
uh suggest something that you could have
7:22
done at the beginning
7:24
so for example maybe you're working with
7:26
something related to
7:28
sound intensity measurements then you
7:30
can you may want to suggest oh i could
7:32
have put
7:33
the sample material to enclose it
7:36
but then again this is something that
7:38
you could have done at the beginning so
7:40
i would say
7:41
these may not look good for you
7:44
so it will be the best if you can
7:47
suggest
7:48
something else or you should explain
7:51
why you couldn't do it at the first time
7:56
second if you look at the criteria uh
7:58
they want you to put
8:00
realistic extension or improvement so
8:03
that is to say
8:04
uh you should not be saying something
8:05
like oh i could use a battery without
8:09
any internal resistance i mean this is
8:12
no way
8:12
it could happen in real life or you can
8:15
say
8:15
simply say a more precise equipment
8:17
again
8:19
it is possible but then it may not be
8:22
realistic
8:22
for the lemon and also the effect may
8:25
not be
8:26
that significant you may also
8:29
think about performing the experiment in
8:31
the
8:32
vacuum space again this is not realistic
8:35
in a secondary high school lab
8:37
remember you should still remain this as
8:40
a secondary school
8:41
lab setting so you can't say i go to
8:44
nasa
8:44
i fancy lab with the equipment to do so
8:47
that is not realistic number three
8:50
that's something that people always
8:52
overlook remember
8:54
that you cannot avoid absolute error so
8:56
that is to say when you use
8:58
a certain measurement tool like ruler
9:00
electronic balance whatsoever
9:02
even with a digital data logger there
9:05
must be
9:06
absolute error involved however you
9:09
cannot avoid that
9:11
but you could reduce the percentage
9:14
error
9:14
okay so the idea is uh thinking about
9:17
the things that you are measuring
9:19
uh you you can calculate the percentage
9:22
error by
9:24
the absolute uncertainty of that over
9:26
the value
9:27
of it right so as long as you increase
9:31
the value of the average then the
9:35
whole thing would decrease the
9:37
percentage error could reduce
9:38
so that is to say maybe you are
9:40
upscaling
9:42
your experiment to a certain size
9:45
then you can increase x log here and
9:48
even though with the same equipment the
9:50
same delta x
9:51
you can still reduce the percentage
9:54
error and that is of course good for
9:55
your experiment
9:57
and that may also relate to number four
10:00
here
10:00
that is uh changing the choice or
10:03
magnitude
10:04
of the control variables so for example
10:07
if say in your framework the equation is
10:12
a equals to b times c
10:15
over d and you are just studying uh
10:18
how the effects of c affecting a uh what
10:21
actually
10:22
happened is you may find out a is very
10:24
hard to measure because it's so small
10:27
comparing to the uncertainty of the tool
10:31
that you are measuring
10:32
so what you can do is you can increase
10:35
b or decrease d so that
10:38
eventually i mean of course these are
10:40
still considerable don't don't fall
10:42
don't get me wrong all right these are
10:43
still considerable it's
10:44
still a constant throughout the
10:46
experiment however if you do this then
10:49
the value the magnitude of a will
10:52
increase
10:53
when you when you do throughout the
10:55
whole investigation so these are
10:57
the idea why we need to have the pilot
11:00
study
11:01
and of course you can further improve it
11:04
to reduce the percentage error of your
11:07
measurement
11:08
by doing this however again
11:12
you may want to explain why you could
11:14
not
11:15
increase this or decrease this or
11:17
whatsoever
11:18
at the beginning of your experiment so
11:20
maybe it's because of the
11:22
reason considering the cause or the
11:25
feasibility or
11:26
the ways of resources etc
11:29
and last two may or may not be
11:31
applicable to your ias
11:34
you may try to check and see whether
11:36
whether or not
11:37
making your independent variable range
11:40
wider will be helpful especially when
11:43
you see
11:44
maybe in your uh data analysis you find
11:47
out the
11:48
relationship is not perfectly confirmed
11:52
or you are not very confident with the
11:55
relationship that you
11:56
establish or there is disagreement with
11:59
your experimental data and the
12:02
literature data
12:03
so uh maybe making it wider or trying to
12:07
do
12:07
more intervals in those areas may
12:10
help you to further investigate
12:14
the validity of the data
12:17
and lastly in some again in some ia
12:20
research question
12:22
you may find there's a loss of energy
12:25
or loss of temperature etc uh for
12:29
some of those research questions maybe
12:32
you can think about whether
12:33
doing it faster so let's just say let's
12:36
say you do it
12:37
in a normal pace then there will be loss
12:40
of
12:41
charge or loss of energy throughout the
12:44
steps
12:45
but then maybe if you can do it faster
12:47
or even think about changing with a
12:49
certain methodology
12:51
you can reduce that effect and that will
12:54
hopefully improve
12:55
your measurement in terms of independent
12:57
variable
12:59
so these are the things that you can
13:01
think about for your improvement and
13:03
future extension you could if this does
13:06
not belongs to any of the error you can
13:08
even set up a new paragraph after the
13:10
table
13:11
to suggest that that's also fine as well
13:14
and again don't be limited by the list i
13:17
gave you here you can always think of
13:18
something else also
English (auto-generated)