ALTERNATIVES TO EXPERIMENTATION because of the manipulated variable.
In other words, dependent variables (DV) is/are the
MODULE FOR CHAPTER 5 outcome/s we are trying to investigate in an experiment. (Myers & Hansen, 2012) In our
example, we see that the variable directly affected by the IV is the participants’ weight.
Therefore, we know that because the experimenters manipulated the sleep hours of the
Overview
participants then measured their weight, we understand that the hours of sleep is the
The previous chapter has taught you about formulating a good experimental hypothesis. It is identified IV and the weight is the DV.
now time to guide you in an in-depth discussion on the Basic Elements of Experimentation.
EXPERIMENTAL OPERATIONAL DEFINITIONS
You will get to explore different important concepts essential in the proposal and conduct of
psychology experiments. The explanation/ elaboration of the meaning of independent variables is known as the
“experimental operational definition.” It specifically shows/ explains what was done to
Topic Presentation
create the various treatment conditions of the experiment. Based on the example earlier, we
In the previous lecture discussions, we have established the definition of a Psychological know that the different levels of the IV was exhibited as the experimenter made use of the
experiment. Now, we once again define the three main features: conditions: very deprived, deprived, and not deprived with the assigned hours of sleep
deprivation respectively.
(1) the manipulation of the antecedent conditions to create at least two different treatment
conditions so that we can evaluate the impact of different sets of antecedents. On the other hand, the dependent variable of an experiment is defined by measured
(2) The exposure of subjects to different treatment conditions so that we can measure the operational definitions. (Myers & Hansen, 2012) It is the description of exactly how a variable
effects of those conditions on behavior. in an experiment is measured. In the example that was used earlier, we say that the weight
(3) The recording of responses or behaviors of subjects under various conditions and of the participants was measured through a weighing scale. Such information is important to
comparing them using statistics. allow other researchers to replicate the steps performed by the experimenters.
INDEPENDENT AND DEPENDENT VARIABLES DEFINING CONSTRUCTS OPERATIONALLY
Independent variables are also known as the dimension that the experimenter intentionally In research, we often use hypothetical constructs/ concepts to explain unseen processes,
manipulates; it is the antecedent the experimenter chooses to vary. For example, the such as hunger, intelligence, or learning, to explain observable behavior. (Myers & Hansen,
experimenter intended to know the effects of hours of sleep on weight. 2012)
For 2 weeks, the experimenters observed three groups of individuals who were intentionally DEFINING NONCONSTRUCT VARIABLES
deprived of sleep (based on their assigned groupings), then, the participant’s weight shall be
recorded before and after the sleep deprivation. We note that their grouping table might On the same account, it is also important to be able to define non construct variables. If an
look like this: experimenter says that their independent variable has levels of sleep deprivation (very
deprived, deprived, and not deprived), it must be operationally defined by saying how many
Very Deprived Deprived Not deprived hours of sleep would It take for an individual to say that he/ she is very deprived, deprived,
1-3hrs of sleep/day 4-6hrs sleep/day 7hrs and up sleep/day and not deprived of sleep. The definition of the conditions are important for the uniformity
Maria Pau Hazel and standardization of the variables being controlled in the experiment.
Ana Kath Amor
Christian Eme Dyne DEFINING SCALES OF MEASUREMENT
Iris Renz Mary
*These are hypothetical data. We recall that we have discussed the Levels of Measurement (the kind of scale used to
measure a variable) in the previous chapter. Hence, we could recall that statistical
With this in mind, what is the Independent Variable (IV)? The independent variable is the treatments to make sense of data depends on the scale of measurement the data will fall
hours of sleep since this is the variable that “varies.” Meanwhile, we also introduce the term into. We have the NOIR:
“Levels” of Independent Variable which means that the IV being manipulated by the
experimenter has two or more values. In the example given earlier, the Independent Variable • Nominal- Classifies items into two or more distinct categories that can be named.
• Ordinal - The magnitude of each value is measured in the form of ranks.
is sleep while the levels are the hours of sleep deprivation as identified beforehand (very
• Interval - Measures magnitude or quantitative size using measures with equal intervals
deprived, deprived, and not deprived). In contrast to an IV, an experiment always has a
between the values.
dependent variable which is defined as the particular behavior we expect to change or vary
• Ratio - Measures magnitude or quantitative size using measures with equal intervals check a variable’s validity, experimenters conduct an assessment to determine whether the
between all values and a true zero point. independent variable was manipulated successfully. This process is known as manipulation
checking.
EVALUATING OPERATIONAL DEFINITIONS
There are three specific types of validity: (a) content validity, (b) criterion related validity,
Why is it important to evaluate operational definitions when conducting a psychological and (c) construct validity.
experiment? With these definitions well presented in a research, an objective evaluation of
readers and critiques could be done, making the replication of a study more feasible. CONTENT VALIDITY is the degree to which the content of a measure reflects the content of
Generally, operational definitions are evaluation by looking at two aspects, their Reliability what is being measured. In simplest terms, it is a measure of validity based on an evaluation
and Validity. of the subjects, topics, or content covered by the experiment. Then, we also consider the
face validity which is defined as the degree to which a manipulation or measurement
RELIABILITY pertains to the consistency and dependability of experimental procedures and Validityis high when a test appears to measure what it is
technique is self-evident. Face validity
measurements. An operational definition is good when it is considered to be reliable in the supposed to measure. This however does not necessarily indicate true validity.
sense that when another experimenter applies it in another experiment, it works in a similar
manner each time.
1. INTERRATER RELIABILITY is the degree of agreement among different observers Criterion- Construct
Content
or raters. Typically used in “content analysis.’ For example, as you were seeking Related Validity
Validity
your professors’ opinion on your instrument to measure online learning, Ma’am Validity (Discriminant
Dyne, Ma’am Iris and Sir Eme expressed that your instrument is acceptable. Then Validity)
we say that your instrument has “high interrater reliability.” This is usually done
through statistical comparisons (Like correlations.)
2. TEST- RETEST RELIABILITY is also known as “temporal stability” and is defined as
the consistency between an individual’s scores on the same test taken at two or Concurrent Predictive Convergent Divergent
more different times. Validity Validity Validity Validity
3. INTERITEM RELIABILITY is the degree to which different parts of a questionnaire,
test, or other instruments designed to assess the same variable achieve consistent
results. (Myers & Hansen, 2012) For example, a researcher devised a 20-item test
to measure mathematical ability of senior high school students in a specific school.
Myers & Hansen (2012) discussed two techniques to measure inter item reliability:
a. Split half reliability involves splitting the test into two halves at CRITERION-RELATED VALIDITY
random then computing for the reliability coefficient between scores
obtained in the two halves. For our example, we may divide the 20- item Predictive validity is the degree to which a measuring instrument yields information allowing
questionnaire into two until there are 10 items for each of the half. Then, prediction of actual behavior or performance while concurrent validity is the degree to
scores for each half will be computed and correlated with each other. which scores on the measuring instrument correlate with another known standard for
measuring the variable being studied.
b. Cronbach’s α (the symbol is read as alpha) is a statistical method to
compute the reliability which considers the correlation of each test item CONSTRUCT VALIDITY is the degree to which an operational definition accurately represents
with every other item. In the mathematical ability test, we compute the the construct it is intended to manipulate or measure. It is said to be the most important
score for each item and correlate it with all the other items. aspect of validity. Thus, construct validity may be elaborated through Convergent Validity
and Discriminant Validity. Let’s say that a group of researchers are interested in looking at
the effects of sleep deprivation on anxiety levels. It is important to put into consideration
VALIDITY refers to the soundness of an operational definition. In experimentation, it refers to that their instrument measures “anxiety” the way we wanted it to be measured and that it is
the principle of actually studying the variables intended to be manipulated or measured. To not measuring other constructs such as their intelligence and political views. On the other
hand, the test that you have created should correlate with other tests that measure anxiety 7. Mortality. The loss of subjects in the various treatment conditions may alter the
levels— convergent validity. Inversely, scores from that instrument shouldn’t correlate with distribution of subject characteristics across the treatment groups.
tests that measure a different construct unrelated to it. 8. Interactions with selection. Some of the foregoing threats to internal validity may
interact with selection to produce effects that are confounded with or
EVALUATING THE EXPERIMENT: INTERNAL VALIDITY indistinguishable from treatment effects. Among these are selection history effects
Internal validity is certainty that the changes in the dependent variable behavior were and selection maturation effects. For example, selection maturation effects occur
actually caused by the imposed/ manipulated independent variable. It is seen to be one of when subjects with different maturation schedules are assigned to different
the most important concepts in experimentation. (Myers & Hansen, 2012) treatment levels.
EXTRANEOUS VARIABLES x CONFOUNDING PLANNING THE METHOD SECTION
A variable other than an independent or dependent variable; a variable that is not the focus The Method section of the Experimental Report includes the description of the experiment
of an experiment but can produce effects on the dependent variable is known as extraneous that was conducted. The details should be sufficient and on point that it may be easily
variables. On the instance that a researcher is experimenting the impact of lighting replicated by future researchers.
conditions on the performance of the participants in an English subject based test. It could be The Methods section has three parts:
inferred that aside from lighting conditions, there are many other variables that could impact
the dependent variable such as the noise and temperature of the room during the 1. Participants/ Subjects. This includes a comprehensive description of the sample included
experiment. An extraneous variable becomes a confounding variable when its value changes in your study. This is important for researchers/ readers to define the external validity of the
systematically along with the independent variable in an experiment. (This will be study conducted.
accentuated in an in-depth manner as we go through chapter 6). Furthermore, a confounding
variable threatens the experiment’s internal validity. As Myers & Hansen (2012) emphasized, 2. Materials. This part provides a very detailed description of the materials that were used
when there is confounding, experimental results cannot be interpreted with certainty and during the conduct of the study. This includes equipment (if special, include the brand,
causal relationships between the IV and the DV cannot be inferred. manufacturer, model, etc.), films, photos, gadgets, and the like that were utilized during the
experimentation... even scripts for instructions that were provided to the participants should
CLASSIC THREATS TO INTERNAL VALIDITY (Donald Campbell) be included in this section.
The following are the classical threats to Internal validity as discussed by Kirk (2014) 3. Procedures. This includes the step by step procedures executed by the experimenter
before, during, and after the experiment.
in his book Experimental Design: Procedures for the Behavioral Sciences.
1. History. Events other than the administration of a treatment level that occur
between the time a subject is assigned to the treatment level and the time the
dependent variable is measured may affect the dependent variable.
2. Maturation. Processes not related to the administration of a treatment level that
occur within subjects simply as a function of the passage of time (growing older,
stronger, larger, more experienced, etc.) may affect the dependent variable.
3. Testing. Repeated testing of subjects may result in familiarity with the testing
situation or acquisition of information that can affect the dependent variable.
4. Instrumentation. Changes in the calibration of a measuring instrument, shifts in the
criteria used by observers and scorers, or unequal intervals in different ranges of a
measuring instrument can affect the measurement of the dependent variable.
5. Statistical regression. When the measurement of the dependent variable is not
perfectly reliable, there is a tendency for extreme scores to regress or move toward
the mean.
6. Selection. Differences among the dependent-variable means may reflect prior
differences among the subjects assigned to the various levels of the independent
variable.
