PSYCHOLOGICAL TESTING

2. Introduction to
Psychological Testing (Part 1)

Psychology is a very diverse field of study

Neuroscienc
e
Biological
Cognitive
Clinical

Social

Developmenta

Psychology is a very diverse field of study united by.

Neuroscienc
e
Psychological
Biological
Cognitive
Testing
Clinical

Social

Developmenta

Why do we need Psychological

Tests?
Provides a set of criteria to help us
Ask more specific questions (Hypothesis Creation)
Systematically collect data to answer those questions
(Hypothesis Testing)
Draw conclusions based on test scores
Continue to scrutinise those conclusions in the face of new
data or new data collection methods

Why is this important?

Empirical studies are the currency of psychological
investigation.
Such studies provide the evidence used to create
and modify psychological theory.
Almost all we know about psychological phenomena
is derived from such studies. So, the content of this
(and all other) psychology courses grounded in
empirical investigation (ie: research studies
employing the scientific method to get answers to
the questions they pose)

Why is this important?

Any discussions and debates about these
theories, including any statements and
arguments you may put forward in your
assignments should be based on the findings of
such studies.

Personal opinions & anecdotal evidence do not

constitute a sound basis for such discussions.

An Example
Your friend, Rob, tells you that:
Watching violence on TV leads to violent
behaviour

Should we be inclined to agree or

disagree with Robs pearls of wisdom?

An Example
Two Choices
1.Ask others what they think
2.Make our own observations
3.Discuss the advantages and limitations of both
these approaches with the person sitting next to
you.

1. Do you think watching

violence on TV causes
aggressive behaviour?
YES 65%

NO 35%

1. Do you think watching

violence on TV causes
This tells us about
aggressive
behaviour?
opinions - not
about the
underlying facts

We are no closer to
discovering
whether TV
violence actually
causes aggressive
behaviour

YES 65%
NO 35%

We need to formulate a
testable hypothesis (or
Watching
prediction)

violence on
TV causes
violent
behaviour

We need to formulate a
testable hypothesis (or
But, what do we
prediction)
mean by:

watching?
violence on TV?
violent behaviour?

We need to formulate a
testable hypothesis (or
Attending to acts of
prediction)
violence screened as part
of TV film, news & light
entertainment
programming causes
individuals to perpetrate
indiscriminate acts of
violence against other
individuals or property

Why do we need Psychological

Tests?
Provides a set of criteria to help us
Ask more specific questions (Hypothesis Creation)
Systematically gather the data to answer those questions
(Hypothesis Testing)
Draw conclusions about our original questions based on the
data we have collected
Continue to scrutinise those conclusions in the face of new
data or new data collection methods

How do we Test this?

Violent Acts
THE DEPENDENT VARIABLE

How do we Test this?

How will we decide what kind of
behaviours constitute violent acts?
Intimidating behaviours?
Acts that cause physical damage?
Acts that cause injury to self/others?
Frequency of behaviour?
Severity of behaviour?

We need to agree on a set of clearly

defined criteria

How do we test this?

How will we decide when
our criteria have been met?
The experimenter could use a
checklist (but beware of bias)
are the children fighting or
playing?
It is better to use an independent
panel of observers and to check
for interrater reliability

Testing Conditions
Exposure to violence on TV
vs
No exposure to violence on TV
THE INDEPENDENT VARIABLE

Testing Conditions
Whats the best way to
achieve the desired
manipulation of the
independent variable?
We could show the Terminator
Trilogy to 25 people on Monday
then observe their behaviour on
Tuesday; next, we could show
the same group Postman Pat
(cartoon) on Wednesday and
observe their behaviour on
Thursday.

Testing Conditions
Whats the best way to
achieve the desired
manipulation of the
independent variable?

We could show the Terminator

Trilogy to 25 people on Monday
then observe their behaviour on
Tuesday; next, we could show
the same group Postman Pat on
Wednesday and observe their
behaviour on Thursday.

Testing Conditions
Better to use 2 groups:
Experimental Group (gets
to watch violent films)
Control Group (gets to
watch non-violent films)
Then observe behaviour of
all individuals from both
groups

EX
PE
RI
G
RO ME
U NT
P
AL

L
O
R
T UP
N O
C O GR

Testing Sample
What kinds of people do we
want to observe?
If we want to be able to
answer questions about
behaviour in general then
we need to observe the
behaviour of the Average
Joe & the Average Joanna

Testing Sample
In reality the best way to
achieve this is by selecting
participants to represent as
broad a cross-section of
society as possible
Care must be taken during
the selection process to
achieve this

Testing Conditions
Similarly, care must be taken
when allocating participants
to the experimental and
control groups
We would risk skewing our
results if we inadvertently
created an experimental
group composed of a specific
group and a control group
composed of an entirely
different group.

EXPERIMENTAL
GROUP

CONTROL
GROUP

Testing Conditions
Ideally, both groups should be
treated in (almost) exactly the
same way, for example:
Both should watch TV for the
same duration
Both should be subject to the
same physical and social
environment
ie: any factor which may
introduce unwanted variability
should be controlled for

Why do we need Psychological

Tests?
Provides a set of criteria to help us
Ask more specific questions (Hypothesis Creation)
Systematically gather the data to answer those
questions (Hypothesis Testing)
Draw conclusions about our original questions based on the
data we have collected
Continue to scrutinise those conclusions in the face of new
data or new data collection methods

Drawing Conclusions From

Psychological Tests
Suppose we have observed:
43 instances of violent behaviour in our
control group (no exposure to violent TV
68 instances of such behaviour in our
experimental group (exposure to violent
TV)
Is this sufficient grounds for claiming that
exposure to TV violence causes violent
behaviour?

Statistical Analyses
We can apply statistical
analyses to the data to
determine whether the
differences observed between
groups are likely to have
happened by chance alone.
If not, then we can suggest
that something else is at work
in producing different amounts
of violent behaviour between
groups.
NB: This is still not grounds
for claiming that violence
on TV causes violent
behaviour.

Making Claims vs Giving Proof

The best we can do is to claim
that our data strongly supports
the notion that TV violence
leads to violent behaviour in
others
In support of such a claim, we
should agree to independent
scrutiny of our experimental
methods and of the data we
have collected

Why do we need Psychological

Tests?
Provides a set of criteria to help us
Ask more specific questions (Hypothesis Creation)
Systematically gather the data to answer those
questions (Hypothesis Testing)
Draw conclusions about our original questions based
on the data we have collected
Continue to scrutinise those conclusions in the face of new
data or new data collection methods

Scrutinising Our Claims (1)

In practice, this means

submitting our results and a
precise account of our methods
to an academic journal

Once submitted, every detail of

our study will be examined by
experts in the field (peer review
by other academics)
Only when they are satisfied
with our account will the article
actually be published and made
available to the wider academic
community.

Scrutinising Our Claims (2)

Once published, our findings
come under wider scrutiny from
researchers and other
interested parties worldwide.
Maybe, Dr X replicates our
study and fails to reproduce our
results (thats bad for our
position)

Scrutinising Our Claims (2)

Once published, our findings
come under wider scrutiny from
researchers and other
interested parties worldwide.
Maybe, Dr X replicates our
study and fails to reproduce our
results (thats bad for our
position)
Maybe, Prof Y runs a variant of
our study using new techniques
and finds an even stronger
effect (thats good for our
position)

Why do we need Psychological

Tests?
Provides a set of criteria to help us
Ask more specific questions (Hypothesis Creation)
Systematically gather the data to answer those
questions (Hypothesis Testing)
Draw conclusions about our original questions based
on the data we have collected
Continue to scrutinise those conclusions in the face
of new data or new data collection methods

Benefits of Testing
We can be more certain that the psychological
effects we are observing are most likely
accounted for by the action of the variables we
are manipulating (the independent variables)
rather than simply happening by chance or
through the action of some unforeseen factors

Benefits of Testing
Throughout this course, we learn more about
How to design a good test
Find out if it is reliable and valid
Learn how to interpret the scores

Definition of Psychological
Test
A psychological test is a systematic
procedure for obtaining samples of
behavior, relevant to cognitive or
affective functioning, and for scoring
and evaluating those samples
according to standards.

Features of Psychological
Test
Samples of behavior
Scoring criteria
Standardization

Features of Psychological
Test
Samples of behavior
Must be representation of the behavior
that is under observation.
Example: pillow fighting children. Is
this a typical violent behavior?
If not, what type of action is
representative of violent behavior?

Features of Psychological
Test
Scoring criteria
Objective: A correct set of answers to a
particular question (eg., SAT tests)
Different experimenters will always
reach the SAME conclusion
Subjective: Relies on experimenters
judgment.
Different experimenters can reach the
DIFFERENT conclusions

Features of Psychological
Test
Standardization
Testing conditions are standardized: they are
the same each time.
Your performance is measured against
someone elses in your age group.

We will take about this topic in more detail

next week.

Some examples of
Psychological Tests
Ability/Achievement Tests: A person
performs a task

Observations: persons behavior is

observed
Self report Tests: A person rates or
reports their feelings, values, attitudes

Examples of Psychological
Tests
Ability/Achievement Tests
Task is well-defined
Individual has to respond
Designed to determine what an individual
CAN do (their ability)

Examples of Psychological
Tests
Ability/Achievement Tests
Task is well-defined
Individual has to respond
Designed to determine what an individual
CAN do (their ability)
Example: SAT, GRE, driving test
Advantage: Objective measure
Limitation: Can be restrictive as there is
no room for creativity.

Examples of Psychological
Tests: Observations
Can be in a lab or a naturalistic setting
(example: one-way mirror)
There is not a well-defined task (unlike tests of
ability)
Example: Infants research
Advantage: Individual may not be aware that
they are being observed so you can gain a good
measure of their true responses.
Limitation: Open to interpretation, different
experimenters have different explanations for
behavior

Examples of Psychological
Tests
Self report/self rating
Relies on the individuals awareness and honesty
Example: Mental health questionnaire
Advantage: The best measure of someones
internal states, because only they are aware of
what they are thinking/feeling
Limitation: People are not always a good judge
of their internal state.
Also, they may want to present themselves in
the best (self serving bias)