RESEARCH METHODS IN
Your course lecturer
• Colin Gray
• Room S16 William Guild Building
• E-mail address: email@example.com
• Telephone: (27) 2233
• Don’t hesitate to get in touch with me if
you have any queries or encounter any
About the course
• Twelve lectures.
• I give my lectures during the first half-session,
but some topics will be useful for the work you
will be doing after Christmas as well.
• The lectures are integrated with other courses
– the practicals
– the workshops on using the computer to analyse data
• A multiple-choice examination.
To help you …
• I shall try to arrange that the overheads for
each lecture are on the Web before the
lecture is given.
• I shall give you plenty of examples of
multiple-choice questions for practice.
• There’s no need for anyone to fail this
course, but I recommend that you
PRACTISE AS YOU GO ALONG.
Purpose of the course
• Psychology is a very wide field of enquiry.
• Different research questions can require
quite different methods.
• General principles of good research
• Psychological research produces
• How to make sense of psychological data
• In these lectures, I shall discuss only the most
elementary statistical methods.
• If you have already taken a statistics course, you
may find some of my later lectures too easy.
• Attendance is not obligatory. Check out each
lecture on the Web first.
• Inspect the exam papers in the library to make
sure that you are adequately prepared.
Kinnear, P. R., & Gray, C. D. (2006).
SPSS 14 made simple. Hove & New
York: Psychology Press.
This book covers most of the topics in this
course of lectures.
It will be useful to you next year if you
continue with psychology.
For a broader view of strategies in
STRATEGY IN PSYCHOLOGICAL
Psychological research is different
from research in a natural science,
such as physics or chemistry
• ‘If you’ve seen one alpha particle, you’ve
seen them all’ (Moore, 1985; p75).
• In physical sciences, the units of study are
• So what is true of ONE alpha particle is
true of ALL alpha particles.
• In chemistry, there are books giving the values
of ‘Physical Constants’.
• In pure form and under specified conditions,
each substance has fixed properties such as a
melting point and a boiling point.
• What’s true of YOUR potassium permanganate
is true of ALL potassium permanganate.
• To discover the properties of your potassium
permanganate (or anyone else’s), just look up
There are no constants in
• Look at the people entering a lecture
• They vary in their appearance, their size
and their physical proportions.
• They vary in their behaviour.
• Their psychological characteristics also vary.
• So do their attitudes and political affiliations.
• Their capabilities also vary.
– How quickly could you slam on the brakes in an
– How well can you catch and throw a ball?
– How quickly and accurately could you identify a guilty
person in a line-up?
• The psychologist must deal with
• A VARIABLE is any characteristic or
property with respect to which objects,
people or situations can differ.
• Gender is a variable. Height is a variable.
The speed with which you can slam on
your breaks in an emergency is a variable.
A key term
Quantitative and qualitative
• Variables such as height, weight and ability level
are characteristics that are possessed IN
DEGREE. They are QUANTITATIVE variables.
• Characteristics such as gender, nationality and
blood group are possessed only IN KIND. They
are QUALITATIVE variables.
• Qualitative variables consist of sets of categories
and are also known as CATEGORICAL
Continuous versus discrete
• Quantitative variables can, in turn, be divided
into those whose values vary CONTINUOUSLY
(height, weight) and those whose values vary in
steps, that is, DISCRETELY.
• The number of visits to a website, for example,
is a discrete quantitative variable.
• So also is the number of heads obtained on ten
tosses of a coin.
Continuous versus discrete …
• A discrete variable can take only a finite
number of values between any two points
on the scale.
• A continuous variable can take an infinite
number of values between any two points
on the scale.
• A MEASUREMENT or DATUM is a numerical
record of a value (or category) of some variable
you are studying.
• A quantitative variable can be measured on an
independent scale with units, such as a ruler, a
weighing machine or a psychological test. This is
‘measurement’ in the everyday sense of the
• I am going to use the word ‘measurement’ in a
more general way to include both quantitative
and qualitative variables.
• Qualitative variables can be recorded by using
arbitrary code numbers, such as 1 for females
and 2 for males. These are also
‘measurements’, in the sense that they are
numerical records of observations.
• Here, however, the numbers are merely acting
as CATEGORY LABELS. The numbers 9 and 99
(for females and males, respectively) would
serve just as well.
• Measurements are also known as DATA.
• DATA (Latin) is the plural form of DATUM.
• Please write, ‘The data ARE interesting’ or
‘These ARE interesting data’, rather than
‘The data IS interesting’ or ‘This IS
• We don’t gather data just for the sake of it.
• In psychological research, the researcher
is often trying to show that some variables
have CAUSAL EFFECTS upon others.
• Does the watching of violent films make
people more violent?
• Do certain drugs improve memory?
• Alzheimer’s disease. Violent pornography
on the internet.
• Research is motivated by a desire to confirm (or refute)
a supposition, or HYPOTHESIS, about nature.
• Often, a hypothesis is a statement to the effect that one
variable affects, or causes, another.
• It might be hypothesised, for example, that learning
speed initially increases with food deprivation but
declines after an optimal level of deprivation has been
• It might be hypothesised that the more violent
programmes that children watch on TV or cinema the
more violent the children actually become.
Three strategies in
Which method is best?
• There is no ‘best’ method. The choice
1. The research question
2. The state of knowledge in the area
3. The circumstances – access to
• Ethical considerations may also
constrain the researcher’s choice of
• Of the three research strategies,
experimentation can yield the strongest
evidence for causality.
• It is much more difficult to establish
causality with the other approaches.
1. Experimental research
• In experimental research, the
experimenter MANIPULATES one variable
to demonstrate that it has a CAUSAL
EFFECT upon another, which is
MEASURED or recorded during the
course of the experiment.
• Such manipulation is the hallmark of a true
The independent and
• The manipulated variable is known as the
INDEPENDENT VARIABLE (IV).
• The measured variable is known as the
DEPENDENT VARIABLE (DV).
• The experimental results should show
whether the IV has a causal effect upon
• Does food deprivation affect speed of
• Participants are tested on learning speed
after different periods of food deprivation.
• Here, the IV (independent variable) is
amount of deprivation and the DV
(dependent variable) is learning speed.
A summary of the results
The true relationship?
The independent variable
• Note that deprivation (the IV) is NOT a
property of the participants in the study – it
is a variable MANIPULATED by the
• Learning speed (the DV), however, IS a
property of the participants and is
measured during the course of the
Direction of causality
• The concept of ‘cause’ raises many problems.
Whole books have been written about it.
• But from the results of this experiment, it is
difficult to escape the conclusion that amount of
deprivation has a lawful effect upon learning
• Initially, the effect of a little deprivation is to
increase speed; but eventually, with increasing
levels of deprivation, speed is impaired.
• The research hypothesis is supported by these
2. Correlational research
• In CORRELATIONAL RESEARCH, all
variables are measured as they occur in
• There is NO MANIPULATED VARIABLE.
• All variables are properties or
characteristics of the people we are
Screen violence and actual
• Does watching violent material promote
actual violence in real life?
• Ethical and practical considerations make
it difficult to manipulate the amount of
violent material that children watch.
• It is easier to measure children on the
amount of screen violence to which they
are exposed (or choose to view) and upon
their actual violence.
A study of violence
• We devise a scale measuring actual
• We devise another scale measuring
amount of exposure to violent
• We measure some children on both
Exposure to screen violence and Actual
• Here is a picture of the
results of our study.
Richard • In this SCATTERPLOT,
each point represents
one of the children.
• Richard got a score of 2
John on Exposure and 4 on
• John got 9 on Exposure
Jim and 8 on Actual.
• Jim got scores of 5 on
both Exposure and
• A statistical ASSOCIATION or CORRELATION
is a tendency for events or values to occur
• If exposure to screen violence promotes actual
violence, we should expect those who watch
more violence to be more violent and those who
watch less to be less violent. We can expect an
ASSOCIATION or CORRELATION between
Exposure and Actual violence.
A strong correlation
• A correlation is evident
from the scatterplot.
• When the shape of a
scatterplot is a narrow
ellipse like this, a strong
correlation is indicated.
• This correlation is about
Scatterplot indicating no
• When the cloud of
points is circular,
there is NO
• We can expect the
correlation to be
Have we really gathered evidence
for the hypothesis?
A famous dictum
does not imply
A causal model
• The hypothesis implies this CAUSAL MODEL.
• The results are CONSISTENT with the
• The correlation may indeed arise because
exposure to violence causes actual violence.
Another causal model
• The child’s violent tendencies towards and
appetite for violence lead to his watching violent
programmes as often as possible.
• This model is also consistent with the data.
Yet another causal model
• NEITHER variable causes the other.
• Both are determined by the behaviour of the
Direction of causality
• Returning to the deprivation experiment, it would
be ridiculous to suggest that learning speed
• In the violence study, however, which was of
CORRELATIONAL, rather than
EXPERIMENTAL design, the direction of
causation is uncertain.
• Indeed, at least three possible MODELS OF
CAUSATION are consistent with the results.
Characteristics of an experiment
• The deprivation experiment embodies two essential
features of a true experiment.
• To show that one variable influences another, we must
have a point of reference. Here we can compare
‘baseline’ performance with performance under varying
degrees of deprivation.
• We must also be able to ‘control’ the experimental
conditions effectively. In this case, the volunteers were
prepared to have their food intake controlled by others
over the period of the experiment.
• You are asked not to
READ the words, but to
NAME the COLOUR of
the print each time.
• It’s difficult to do, because
the natural tendency is to
read the word.
• The more words there
are, the more difficult it
• The Stroop task is obviously difficult.
• But you have not participated in a true
• There is no element of comparison here: there
was only one condition.
• We need to compare the task you have just
attempted with the task of naming the same
colours in an array similar to that used for the
• We need a comparison or CONTROL condition.
The rule of one variable in
• In a true experiment, there must be at least two
• If there are just two conditions, one of them is usually a
comparison or CONTROL condition.
• The comparison or CONTROL condition must differ from
the active or EXPERIMENTAL condition only in the
absence of the active agent. In all other respects, the
two conditions must be identical.
• In the Stroop experiment, participants must also be
asked to name the colours of non-word patterns.
A control (comparison) condition
• In fact, it is much
more difficult to name
colours than to read
• Even so, Stroop
(1935) showed that it
takes much longer to
name the print colour
of the conflicting
Extraneous or confounding
• It is often very difficult to devise an
appropriate control condition in
accordance with the rule of one variable.
• If the control condition differs from the
experimental condition in more than one
way, this characteristic may be entangled
or CONFOUNDED with the independent
3. Observational research
• In the third kind of research, the
researcher neither manipulates variables
nor measures them as they occur in
• In OBSERVATIONAL RESEARCH, the
researcher simply records what the
participant says and does.
• As examples, we have the classic
observational studies of animal behaviour.
• In his book King Solomon’s Ring, the
naturalist Konrad Lorenz describes his
highly influential observations on such
behaviour as imprinting in ducklings.
• The French entomologist Jean-Henri
Fabre observed and described the
behaviour of the hunting wasp.
• In psychology or psychiatry, the
counterpart of the natural history study is
the CASE HISTORY, the reports of
Sigmund Freud and Kraftt-Ebbing being
The participant observer
• What causes football violence?
• Researchers have infiltrated football gangs and
attempted to discover the social dynamics of violence.
• Big business makes extensive use of techniques of
persuasion, and not just in advertising. The Tupperware
organisation trained a large staff to organise parties at
which those present were persuaded to host their own
parties later. Specially trained Tupperware personnel
would bring great pressure to bear upon guests to make
a commitment to host their own parties at some specified
• The Tupperware training methods have been described
by Taylor (1975), who arranged for members of his
research team to infiltrate a Tupperware group and
observe their methods.
Phenomenology: Enquire within
• PHENOMENOLOGY (Brentano) is a
philosophical doctrine which held that
psychology should be based upon
• There have always been psychologists
who argue that psychology should be less
about numerical data and more about
people’s descriptions of how things seem
to them personally.
• The principal research tool of the
phenomenologist is INTROSPECTION, which is
description of one’s own thoughts and
• This method is still used in modern psychology,
as when chess players and other problem-
solvers are asked to ‘think out loud’ as they play.
• It is hoped that this INTERIOR MONOLOGUE
will reveal the nature of the thought processes
Advantages of observational research
• Experiments can seem highly artificial.
• Observational studies in natural settings
seem to have more ECOLOGICAL
• Modern advances in technology have
provided powerful aids to observational
work. Infra-red photography, for example,
has contributed enormously to our
knowledge of nocturnal animal behaviour.
• We now know much more about the true habits
of hyenas, once thought to be cowardly
scavengers, as they roam the noctural African
• Similarly, CCTV footage has revealed the
behaviour of packs of human predators as they
roam a nocturnal urban landscape as
nightmarish as that envisaged by Anthony
Burgess in his book Clockwork Orange. His
dystopic vision has proved to be prophetic.
Burgess’s creation, Little Alex, who horrified a
generation, is alive and well today.
The qualitative approach
• Recent years have seen a resurgence of
phenomenology in the form of
QUALITATIVE RESEARCH (Rennie,
Watson & Monteiro, 2002).
• Buchbinder & Eiskovits (2002), having
studied the detailed accounts of 20
women’s experiences of domestic violence,
concluded that it is primarily shame that
prevents victims of domestic violence from
reporting the crime to the police. 63
Problems with observational
• The observer is not like a video-recorder any
more than memory is like a filing cabinet.
• Like memory, observation is a highly selective
process and driven by the assumptions and
expectations of the observer.
• There are numerous records of eminent
scientists seeing what they expected to see,
rather than what was actually there.
• Seeing is believing; but believing can also
lead to seeing and what is seen may not
actually be there.
• The microscopist Von Hartsoeker, a
preformationist, believed that he could
see, in a human spermatozoon, a tiny
baby in such clarity of detail that he was
able to draw it.
Introspection – or retrospection?
• Arguably, INTROSPECTION is actually
RETROSPECTION and hindsight can be
• Critics of the ‘thinking aloud’ approach to
chess have argued that such verbalisation
not only fails to reveal the true processes
involved, but may actually alter the
thinking of the solver.
• The process of any kind of observation, whether in a
naturalistic or a laboratory setting, can affect the nature
of what is measured.
• The participant observer who has infiltrated a football
gang may, in order to conceal his true role and hidden
agenda, feel forced to influence the social dynamics of
• The manner in which he does this may reflect his own
ideas about the dynamics of group violence.
• The Buchbinder and Eiskovits study raises several
questions. Is the explanation of the women’s reticence
really shame, as they say, or something else?
• Did those authors read their own preconceptions into the
• Much psychological research attempts to
show that some variables exert a causal
influence over others.
• I reviewed three main strategies:
• Each approach has its advantages and
• quantitative versus qualitative variables.
• continuous versus discrete variables.
• measurement, datum, data.
• experimental, correlation and observational
• Independent variable, dependent variable.
• correlation versus causation
• causal model
• Stroop interference
• demonstrations versus experiments
• rule of one variable
• extraneous variables, confounding
Key terms …
• case history
• participant observation
• interior monologue
• ecological validity
• qualitative research
Key terms …
• belief-motivated perception
• Instrument reactivity