Research Methods in Psychology ~Siddhant Jhawar

1. Experiments

An experiment is an investigation where psychologists look for a causal relationship

between the independent variable (IV) and the dependent variable (DV), such as an
investigation looking to study a causal impact of differing brightness of light on labor
productivity.

Independent Variable (IV): The factor under investigation that is manipulated by the
psychologists to create two or more experimental conditions (levels of IV) and is
expected to result in a causal change in the dependent variable (DV), such as the
differing brightness of light to test its effect on labor productivity.

Dependent Variable (DV): The factor in an experiment that is measured by the

psychologist and is expected to change causally under the influence of the
independent variable (IV), such as the differing measured levels of labor productivity
influenced from the manipulated brightness of light.

Experimental Conditions: One or more of the situations in an experiment that

represent different levels of the IV and are compared to each other or to a control
condition.

Control Condition: A level of the IV in an experiment from which the IV itself is absent.
It is compared to one or more experimental conditions. For instance, the normal
brightness of the light in the factory is the control condition since the brightness has
not been manipulated the IV is absent from this situation.

However, almost always, there are differences in the DV that have resulted not from the
manipulation of the IV. These causal differences in the DV result from what are known
as uncontrolled variables.

Extraneous (Uncontrolled) Variable (EV): A variable which acts randomly affecting the
DV in either all or, systematically, any one experimental condition (level of IV). This can
obscure or confound the effect of the IV, making the causal relationship between the IV
and the DV even more difficult to interpret. For instance, some workers already have a
higher productivity than other workers in a factory, so despite changes in the light
brightness, these workers may continue to produce at a higher productivity.

Confounding Variable: A form of EV that acts systematically on one level of the IV, so
it could hide or exaggerate differences between the experimental conditions and

Page 1

Research Methods in Psychology ~Siddhant Jhawar

therefore confound the results making it difficult to interpret the causal relationship
between the IV and the DV.

Confound means Confuse.

Participant Variable: A form of EV where individual differences between participants

(Ps), such as their age, intelligence quotient, and personality, affect their behavior in a
study, resulting in either underpinning or exaggerating the differences in the DV
between the experimental conditions.

Situational Variable: A form of EV where the situation, or the environment, can affect
the DV rather than the IV. Common examples of situational variables include noise,
temperature, lighting, and weather conditions.

Page 2

Research Methods in Psychology ~Siddhant Jhawar

2. Experimental Designs

An experimental design is the way in which participants (Ps) are allocated to levels of
the IV (experimental conditions).

Independent Measures Design: A form of experimental design in which a different

group of Ps is used for each level of the IV (condition).

Strengths of an Independent Measures Design Weaknesses of an Independent Measures

Design

As Ps only take part in one experimental condition This design does not always adequately control
they are less likely to respond to clues that result for participant variables (EV), such as individual
in them guessing the aim of the experiment, differences between the Ps. This suggests that
therefore reducing the potential effects of even by chance, all Ps of a certain trait or
demand characteristics and raising the validity of personality might be exposed to one
the study. experimental condition and all Ps of a different
trait or personality might be exposed to the other
condition. This means that there remains a
possibility that individual differences, rather than
the IV, is affecting the DV.

However, this risk can be reduced by random

allocation of Ps to different conditions. This
spreads possible differences between individuals
across the levels of the IV.

As Ps only take part in one experimental condition More Ps are needed than in a repeated measures
there are no order effects that can reduce the design so the study may be less ethical if
validity of the findings. participants are harmed and less effective if there
is a small sample because participants are hard to
find.

Demand Characteristics: Specific features of the experimental situation which give

away the aim of the study. They can cause the ps to try to change their behavior, for
example to match their beliefs about what is supposed to happen, which reduces the
validity of the study.

Hawthorne Effect: The modification of behavior by Ps in response to their knowledge

that they are being observed in a study by the Es.

Random Allocation: A procedure to reduce the effect of confounding variables, such

as individual differences. Ps are put in each level of the IV such that each person has an

Page 3

 Research Methods in Psychology ~Siddhant Jhawar

equal chance of being in any condition. For instance, each P is given a number and the
numbers are then randomly divided into two groups, each being allotted to an
experimental condition. This is possible by using a random number generator on a
computer or picking numbers from a hat.

Repeated Measures Design: A form of experimental design in which each P is

exposed to every level of the IV (condition).

Strengths of a Repeated Measures Design Weaknesses of a Repeated Measures Design

This design adequately controls for participant A major flaw is that the design can create order
variables (EV), such as individual differences effects. If a P performs an activity twice, they may
between the Ps. This is because all Ps take part in become tired or bored the second time, known as
all experimental conditions therefore each P acts the fatigue effect, and the result is different from
as their own baseline. Any individual differences the first time. It might be that the second result is
would influence the DV in the same manner across much better than the first because the P knew
all levels of the IV. Therefore, the findings are less what to expect or treated the first as a practice.
likely to be biased, raising their validity. This is simply known as the practice effect.

However, this risk can be reduced by

counterbalancing. Here, the group of Ps is
divided into two, so one will do condition A
followed by B and the other half will do B
followed by A. Another way to reduce this risk is
by employing randomization. Here, each P is
allocated to perform in the different levels of the
IV in a way that ensures they have an equal
chance of participating in different levels in any
order.

Fewer Ps are needed than in an independent As Ps see the experimental task more than once,
measures design so the study may be more they have greater exposure to demand
ethical if participants are harmed and more characteristics.
effective if there is a small sample because
participants are hard to find.

Order Effects: Order effects are the consequences of participating in a study more
than once, for example in a repeated measures design. They cause changes in
performance of Ps between conditions that are not due to the IV, so can obscure or
confound the effect on the DV.

Page 4

 Research Methods in Psychology ~Siddhant Jhawar

Practice Effects: A situation where Ps’ performance improves because they experience
the experimental task more than once, for example due to familiarity or learning the
task.

Fatigue Effects: A situation where Ps’ performance declines because they experience
the experimental task more than once, eg. Due to physical tiredness or boredom with
the task.

Randomization: A procedure to even out order effects in a repeated measures design.

Each P is allocated to perform in the different levels of the IV in a way that ensures they
have an equal chance of participating in the different levels in any order.

Counterbalancing: A procedure to even out order effects in a repeated measures

design. Each possible order of levels of the IV is performed by a different sub-group of
Ps. This can be described as an ABBA design, as half the Ps do condition A followed by
B, and the other half do B then A.

Matched Pairs Design: A form of experimental design in which Ps are arranged into
pairs. Each pair is similar in ways that are important to the experimenters (Es) and one
member of each pair performs in a different level of the IV.

Strengths of a Matched Pairs Design Weaknesses of a Matched Pairs Design

Ps see only one level of the IV, reducing the effect The similarity between pairs is limited by the
of demand characteristics, raising their validity. matching process, so the right matching criteria
must be chosen in advance for this to be effective.

Participant variables are less likely to distort the Availability of matching pairs may be limited,
effect of the IV than in an independent measures making the sample size small (although some
design as individual differences are matched. studies conducted on twins use very large
numbers of pairs).

Different Ps are used in each level of the IV so

there are no order effects.

Page 5

 Research Methods in Psychology ~Siddhant Jhawar

3. Types of Experiments

Laboratory Experiments: An investigation in which there is an IV, a DV and strict

controls. It looks for a causal relationship and is conducted in a setting that is not in the
usual environment for the participants with regard to the behaviour they are
performing.

Strengths of a Laboratory Experiment Weaknesses of a Laboratory Experiment

The manipulation of one IV while controlling EVs The results may be biased by sampling, demand
means that causal relationship is more likely to be characteristics or experimenter bias.
shown.

They allow for control over many extraneous Some people regard the process as
variables, e.g. temperature and noise levels. dehumanising, with participants being treated like
laboratory rats by having something done to
them.

Standardised procedures and operational Controlling variables is reductionist as it is

definitions (operationalizations) mean that unlikely that any behaviour would exist in isolation
replication is possible. from other behaviours.

In a laboratory, participants must have given some Artificial conditions (setting and task) can produce
degree of consent, but not necessarily informed unnatural behaviour, which means that the
consent, to take part. research lacks ecological validity. The behaviour
is like to be unnatural, and so it does not tell us
how people behave in real life. Hence, the results
suffer from low mundane realism. Thus, the
findings are not easily generalizable to actual
settings.

For the IV to be isolated, participants might be

deceived about the true nature of the study. There
may be other ethical issues.

It is more likely that the data will be snapshot

rather than longitudinal.

Page 6

 Research Methods in Psychology ~Siddhant Jhawar

Field Experiments: An investigation looking for a causal relationship in which an IV is

manipulated and is expected to be responsible for changes in the DV. It is conducted
in the natural environment for the Ps for the behaviour being investigated.

Strengths of a Field Experiment Weaknesses of a Field Experiment

There is greater ecological validity because the There might be difficulties in controlling the
surroundings are natural. situation (uncontrolled situational variables), and
therefore more possibility of influence from
extraneous variables.

If people are unaware of the research taking The experiment might be difficult to replicate
place, there is less likelihood of demand exactly as the procedure becomes more difficult
characteristics to standardize.

The features of an experiment (IV, DV etc.) are There might be problems of access to where the
retained. study is to be done; such as consent from a
company.

The behaviour is natural and so tells us how There might be ethical issues of consent,
people behave in real life. Hence, the results do deception, invasion of privacy etc.
not suffer from low mundane realism. Thus, the
findings are more easier to generalize to actual
settings.

Controls: Ways to keep potential confounding variables constant, for example

between levels of the IV, to ensure that measured differences in the DV are likely to be
due to the IV, raising validity.

Standardization: Keeping the procedure for each P in a study exactly the same to
ensure that any differences between Ps or conditions are due to the variables under
investigation rather than differences in the way they were treated.

Replication: Keeping the procedure and materials exactly the same between studies
when attempting, for example, to verify results or to enable other studies to use the
same techniques to answer related questions.

Operationalizations: The clear description of a variable such that it can be accurately

manipulated, measured or quantified, and the study can be replicated. This includes
the way that the IV and DV in experiments, and the co-variables in correlations, are
described.

Page 7

 Research Methods in Psychology ~Siddhant Jhawar

Placebo: A pill or procedure given to a patient who believes it to be a real treatment

which in fact has no active ingredient, i.e. no active drugs in the case of a pill or no
therapeutic value in the case of an intervention. (Has no direct relation to experiments
but can be asked to define).

Pilot Study: A small scale test of the procedure of a study before the main study is
conducted. It aims to ensure that the procedure and materials are valid and reliable, so
that they can be adapted if not.

Generalize: To apply findings of a study more widely, e.g. to other settings and
populations.

Ecological Validity: The extent to which the findings of the research conducted in one
situation would generalize to other situations. This is influenced by whether the
situation (e.g. a laboratory) represents the real world effectively and whether the task is
relevant to real life (has mundane realism).

Page 8

 Research Methods in Psychology ~Siddhant Jhawar

4. Reliability, Validity, and Ethics of Experiments

Reliability: The extent to which a procedure, task, or measure is consistent, for

example, that it would produce the same results with the same P on each occasion.

Validity: The extent to which the Es are testing what they claim to be testing.

Internal Validity: Refers to the extent to which the IV is directly affecting the DV and
not some EV.

External Validity: Refers to the extent to which the findings of the study can be
applied to real-life settings and to other people outside the sample.

Page 9

Research Methods in Psychology ~Siddhant Jhawar

5. Hypotheses in Experimental Studies

Hypothesis: A testable statement based on the aims of an investigation. A hypothesis

should be falsifiable.

Null Hypothesis: A testable statement which states that any difference or correlation in
the results is completely due to chance, that is, that no pattern in the results has arisen
because of the variables being studies. For instance, any difference between the labor
productivity and levels of light brightness is due to chance.

Alternative Hypothesis: A testable statement predicting a difference in the DV

between levels of the IV in an experiment or between co-variables in a correlation.

Non-Directional (Two-Tailed) Hypothesis: A form of an alternative hypothesis

predicting only that one variable will be related to another, for example, that there will
be a difference in the DV between levels of the IV in an experiment or between co-
variables in a correlation. For instance, there is a difference between the labor
productivity and levels of light brightness.

Directional (One-Tailed) Hypothesis: A form of an alternative hypothesis predicting

the direction of a relation between variable, for example in an experiment whether the
levels of the IV will produce an increase or a decrease in the DV or whether one co-
variable will produce an increase or decrease in another co-variable. For instance,
workers working in an environment with a brighter light setting will be experiencing a
greater productivity.

6. Ethics of an Experiment

Informed Consent: Knowing enough about a study to decide whether you want to
agree to participate or not.

Right to withdraw: A P should know that they can remove themselves, and their data,
from a study at any time.

Protection from harm: Ps should not be exposed to any greater physical or

psychological risk than they would expect in their day-to-day life.

Deception: Ps should not be deliberately misinformed (lied to) about the aim or
procedure of the study. If this is unavoidable, the study should be planned to minimize

Page 10

 Research Methods in Psychology ~Siddhant Jhawar

the risk or distress, and Ps should be thoroughly debriefed. It may be done to reduce
the effects of demand characteristics but should be avoided.

Privacy: Ps’ emotions and physical space should not be invaded, for example they
should not be observed in situations or places where they would not expect to be
seen.

Con dentiality: Ps’ results and personal information should be safely and not release
to anyone outside the study.

Page 11
fi