SURVEY METHOD ASSIGNMENT

DEWI IRIANI
566751

1. Explain the main advantages of the mixed mode in data collection techniques
If we are talking about statistical data, in most cases we will have to deal with error. In
particular, in conducting a survey there is more likely that we have to face the total survey
error, which is including sampling errors and non-sampling errors ( that is coverage error, non-
response error, measurement error). Mixed mode in data collection offers the possibility of
reducing this errors.
In using mixed-mode data collection we meant data is collected in multiple ways from the
sample in one period survey. The collection mode could use printed papers and electronic
questionnaires (either on-line or off-line) or even by phone called interview. mixed mode data
collection is used to get high response. However, it is not automatically superior to a single
method data collection Likewise, mixed method approaches do not necessarily produce higher
quality data. Mixed method data collection is only likely to produce high quality data if each of
the component parts are of the highest possible quality 1. Bellow I list the disadvantages and
advantages of using the mixed mode collection data technics
Compares quantitative and qualitative data. Mixed methods are especially useful in
understanding contradictions between quantitative results and qualitative findings.
Reflects participants’ point of view. Mixed methods give a voice to study participants and
ensure that study findings are grounded in participants’ experiences.
Fosters scholarly interaction. Such studies add breadth to multidisciplinary team research by
encouraging the interaction of quantitative, qualitative, and mixed methods scholars.
Provides methodological flexibility. Mixed methods have great flexibility and are adaptable to
many study designs, such as observational studies and randomized trials, to elucidate more
information than can be obtained in only quantitative research.
Collects rich, comprehensive data. Mixed methods also mirror the way individuals naturally
collect information—by integrating quantitative and qualitative data. For example, sports
stories frequently integrate quantitative data (scores or number of errors) with qualitative data
(descriptions and images of highlights) to provide a more complete story than either method
would alone.
1) William G, Axinn, Lisa D, Pearce Mixed method data collection strategies

2. What is the difference between Objects and Units in the two-phase quality framework
for statistics based on combined data sources?
We can see the difference between object and units by the following characteristics:
Objects

 Objects: Identify the administrative recording units that maybe different from the
statistical units. Very often the administrative objects are events.
 Concepts: refer to the administrative measures that maybe different from the statistical
target variable
 Set: refers to the administrative population that maybe different from the statistical one

 Coverage Error The difference between the objects in the target population and the
objects in the available administrative set e.g. SSD register only regular jobs

 Non-response Error The difference between the accessible administrative set and the
observed set, where the observed set referred to the one that contains only the
registered objects e.g. not registered regular job events

 Processing Error The difference between the observed set and the edited set, where the
edited set referred to the one that contains the processed record according the
administrative rules e.g. job events inadmissible according to the administrative
regulation Validity Error arise when the target concept, defined according to the
administrative interest, and the measures differ. e.g. SSD concepts are very clear
because they are collected for administrative purpose

 Measurement Error arise between the intended measurement and the obtained
response.

 Processing Error arise between the response and edited response.

Units

Phase two is different from the life cycle of sample surveys, because introduce the integration
phase needed for the realization of the target population and concepts. Here target population
and target concepts referred to statistical population (units) and statistical concept (variables).
Coverage Error The difference between the units in the target population and the units in the
linked sets, where the linked sets is the set of units deriving from the linkage of different
sources. At this stage, coverage errors may derive from the information in the administrative
sources, but they can also occur through linkage errors e.g. SSD register measurements errors
on the activity status may affect the coverage of the linked sets Identification Error Occurs
when we need to identify composite units starting from the base units. Generally, the relation
between the base and the composite units is many to one. Identification error occurs when
there are errors regarding the relation between the base units e.g. composite units are families,
enterprises, enterprise groups Unit Error Once the statistical units have been identified, we
need to validate them e.g. identification of active enterprises

3. Describe the R functions to create histograms and barplots also describing the main
differences between the two kind of plots.
The histogram consists of an x-axis, an y-axis and various bars of different heights. The y-axis
shows how frequently the values on the x-axis occur in the data, while the bars group ranges of
values or continuous categories on the x-axis. The latter explains why histograms don’t have
gaps between the bars.
step one: to make a histogram, we indeed need to some data to be plotted. We can do it by
importing or using that has already built in to R. second step is making a histogram by using the
hist function in which compute histogram ofthegiven data values. by putting the name of our
data set in between the parentheses of this function, like this: for xample (hist(AirPassengers).
However, if you want to select only a certain column of a data frame, chol for example, to make
a histogram, you will have to use the hist() function with the dataset name in combination with
the $ sign, followed by the column name: hist(chol$AGE) #computes a histogram of the data
values in the column AGE of the dataframe named “chol”
The default visualizations usually do not contribute much to the understanding of your
histograms. You therefore need to take one more step to reach a better and easier
understanding of your histograms. While still using the hist() function.
In order to adapt your histogram, you simply need to add more arguments to the hist()
function, just like this:
hist(AirPassengers, main="Histogram for Air Passengers", xlab="Passengers",
border="blue", col="green", xlim=c(100,700), las=1, breaks=5)
This code computes a histogram of the data values from the dataset AirPassengers, gives it
“Histogram for Air Passengers” as title, labels the x-axis as “Passengers”, gives a blue border
and a green color to the bins, while limiting the x-axis from 100 to 700, rotating the values
printed on the y-axis by 1 and changing the bin-width to 5.
Bar charts (also called Bar plots) are useful for the quantitative comparison of a grouped data. A
bar chart represents the data with rectangular bars such that the length of the bar is
proportional to the values of the quantity they represent. The individual categoroes are
represented as vertical or horizontal bars standing next to each other for quantitative
comparison.
Create barplots with the barplot(height) function, where height is a vector or matrix. If height is
a vector, the values determine the heights of the bars in the plot. If height is a matrix and the
option beside=FALSE then each bar of the plot corresponds to a column of height, with the
values in the column giving the heights of stacked “sub-bars”. If height is a matrix and
beside=TRUE, then the values in each column are juxtaposed rather than stacked. Include
option names.arg=(character vector) to label the bars. The option horiz=TRUE to createa a
horizontal barplot.

For example
# Simple Bar Plot
counts <- table(mtcars$gear)
barplot(counts, main="Car Distribution",
xlab="Number of Gears")
Bar plots need not be based on counts or frequencies. You can create bar plots that represent
means, medians, standard deviations, etc. Use the aggregate( ) function and pass the results to
the barplot( ) function.

The Difference Between Bar Charts and Histograms