You are on page 1of 44

|  


 


    



 



`  




  
Learning objectives
1. Exploring research methods in psychology (a
general account)
2. Applying research methods (in the study of
memory)
3. Quantitative and Qualitative research
methods in psychology
4. How do we design a study (in detail)
5. Are we as teachers ͚committed͛ to using
empirical methodology as scientists?
!tarting with a story...
[ rn winter people wear [ Observation
heavy clothes
[ The weather outside is cold [ Rationale
and there are storms
[ People wear heavy clothes
during winter in order to [ Hypothesis
protect themselves
[ rt is so because it is winter
[ Prediction





     




 
 

Examples of research designs
°etween-participants/Within
participants designs
Types of research
åaking sense of our data via...


  
    

 

     






   
 !

Applying research methods in the study of


memory (an example)
!ummary No. 1

[ Research methods are a useful tool in


psychology towards studying human
behaviour
[ Different research types as well as statistical
techniques are employed for different aspects
of research
[ rn designing a study we pose a research
question in order to seek evidence that will
support our hypothesis
Quantitative and Qualitative Research
åethods in Psychology

     
 

[ Advantages/Disadvantages [ Description of each method
of research methods related [ °eing aware of at least two
to the scientific nature of advantages and two
psychology weaknesses
[ Experiments [ How research methods relate
to the scientific nature of
[ rnvestigations through psychology
correlational analysis
[ Distinguishing psychological
[ Naturalistic observations knowledge from
[ Questionnaires commonsense
[ rnterviews [ Ethical guidelines
Terminology (1)
[ Ways in allocating participants to participate to a study: randomly,
opportunity sample (friends, relatives, and anyone who is around that
time), self-selection (via advertising the study, and finding participants
who volunteered themselves ʹ volunteer sample)
[ r (rndependent ariable): rt is manipulated by the experimenter, i.e. the
experimenter puts down the ͚rules͛ to be followed by each participant;
each r is independent of the value of any other variable (r )
[ D (Dependent ariable): rt is depended on the value provided by the r
(existence or dismissal of such dependence is questioned in relation to the
hypothesis investigated)
[ One-tailed/Two-tailed hypotheses (othewise called: directional/non-
directional hypotheses: One-tailed or directional hypothesis refers to the
increase or decrease of one variable in favour to the other; once there
may be an effect of one variable to the other, and the direction of this
effect is unknown we have a two-tailed or non-directional hypothesis
[ Null hypothesis: No significant difference (in our case: no real effects in
measuring the blood pressure of participants)
Terminology (2)
[ °etween participants/within participants: randomly
allocating participants to different groups/no random
allocation and all participants perform on all conditions
[ Probability alue (p): rs normally set to p<0.05 (50-50
chance of finding an effect), or p<0.01 (10% chance of
finding an effect)
[ Effect size (d): The magnitude of the difference
between conditions (the difference between the
means of both or more conditions of a study in terms
of their standard deviations): xІ-xЇ
mean !D
Research methods listing labels (1)
 


 "
 
| #   

Lab experiment åanipulation of the For: Causal Deception;


r to find effects on conclusions; No rnformed consent;
D under confounding Psychological harm
controlled variables;
conditions Replication
Against: Artificial;
Experimenter and
Participant effects
Field experiment Looking for causal For: Causal rnformed consent;
relationships in conclusions; Difficulty in
natural places Ecological validity; debriefing; Privacy
Participants͛ effects
are avoided
Against: Less
controls; Time
consuming
Research methods listing labels (2)
 


 "
 
| #   

Natural experiment Non-manipulation For: Look for Confidentiality


of r ; No random research where r is
allocation of not manipulated;
participants; Quasi- Real problem
experimental studies
Against: No causal
relationships can be
found; åany
confounding
variables
Correlational Co-variables are For: ariables åisunderstanding
analysis studied for positive, cannot be of findings
negative, or zero manipulated;
association Correlation does
not infer causation
Against: Can lead to
misinterpretation;
Other variables can
also come to play
Research methods listing labels (3)
 


 "
 
| #   

Naturalistic ariables are free to For: Can study rnformed consent;


observations vary behaviour where Difficulty in
cannot manipulate debriefing; Privacy
variables; Ecological
validity
Against: No control
of confounding
variables; Observer
bias, low observer
variability
Questionnaires !et of questions For: Collection of Confidentiality;
huge amount of Privacy
data; No specialist
administrators
required
Against: !ocial
desirability bias;
°iased samples
Research methods listing labels (4)
 


 "
 
| #   

rnterviews Pre-determination For: Rich data; Confidentiality;


of questions, or Telephone Privacy
created in response interviews
to answers Against: !ocial
desirability bias;
!killed personnel is
required
Design & rmplementation (1)
[ Assumed study: A psychologist chose to
investigate the kinds of tasks which created stress
for teachers. He asked for teachers from several
schools to volunteer to take part in the study, and
selected 20 male teachers. Each volunteer was
asked to perform two tasks: Presentation of a
prepared speech to a group of 200 students;
marking a set of books full of mistakes. After each
task the researcher recorded the blood pressure
of each participant
Design & rmplementation (2)
[ Research method: Field experiment (Class environment reactions are
tested, both at the speech as well as at the marking tasks)
[ Aim of the study: rnvestigating kinds of tasks to see if they create stress to
teachers
[ Hypothesis: Teaching to a large number of pupils as well as marking their
homework may have an effect in teachers͛ blood pressure. The hypothesis
will be non-directional
[ Design: Quasi-experimental, because we don͛t randomly allocate
participants to conditions; we have only one gender group of participants;
we cannot manipulate each participant in terms of each task performance
[ r /D : Teaching before a large group & åarking homework ʹ 2 r s; °lood
pressure recordings (score noted for each participant) - D
[ Within participants study: All participants will take part in both conditions
[ Co-variables in the study: Teaching before a large group & åarking
homework
Design & rmplementation (3)
[ Pros & cons of the design employed: Participants seem to not having been informed about
the purpose of the study, and then to be asked to take place; no reference to how the
experimenter would deal with the data collected
[ Confounding/extraneous variables: No reference to the age of each participant, so individual
differences in performance to be discussed; anxiety proneness of teachers prior to teaching
was not taken into account, so to possibly influence their performance during the study as
well as their blood pressure scores after both tasks; speaking to public as a stressful life event
was not discussed; female participants could also be employed
[ Ethical issues: Lack of further procedural information to participants before the study to be
taken place, such as anonymity, debriefing, etc
[ rnternal validity of the study: No reference about the subject of the talk to pupils, or if each
participant could choose a subject of his own likeness; if so, whether different topics affect
teachers͛ performance
[ External validity: rt isn͛t known if speaking to the public out of a class environment would
have the same results; external validity is not clear in many types of the experimental designs
[ Ecological validity: A variety of reasons could fall in replicating the study with different
teachers in different classes; probably the study is not ecologically valid, i.e. it cannot rely on
the collection of the same data when applied to different environments
Design & rmplementation (4)
[ Discussion: The study could be replicated by asking all participants to perform on
the same tasks via two different conditions for each of the r s: easy topic/difficult
topic, and good homework/bad homework conditions
[ !election of participants: olunteer sample (self-selection)
[ Pilot study prior the study to be conducted: Half of volunteer teachers prepared a
speech and delivered it to a small audience of students; they also marked the
scripts of both exceptional and non-exceptional students. rn this way, confounding
variables such as the ones named above could be avoided
[ !tatistical procedure possibly to fail: °y what means of an instrument the talk and
marking of homework would be measured? Would that be a quantitative approach
or qualitative? rn which way data will be collected in terms of the performance of
teachers, so these to be discussed in relation to their blood pressure scores after
both tasks had been performed? Also, would increased of decreased blood
pressure be associated to both tasks, or to one of them, or otherwise (looking for
stressful life events other than the participation to these tasks)?
!ummary No. 2
[ Different research methods are used for different
types of studies
[ °y clearly stating the aims, the hypothesis, and
the design of a study we can carry on with the
other issues associated to the study
[ Ethical issues should be presented before
embarking on a study
[ The ways to selecting participants can help us to
identify possible confounding variables related to
the execution of the experiment
How do we design a study (in
detail-1)
[ A research project into eating problems
sought to find out about adolescent and pre-
adolescent dieting practices. The study
involved the use of a questionnaire, including
both quantitative and qualitative questions.
The questionnaire was piloted with a small
group of university students
How do we design a study (in
detail-2)
[ Aspects to be considered in the design of this study
1. Research method
2. Aims
3. r /D
4. Hypothesis
5. 1-tailed or 2-tailed hypothesis
6. !election of participants
7. Quantitative (numbers)/Qualitative (interviews)
8. Possible confounding variables not only to the type of
study, but also to the use of a questionnaire, and the
selection of participants
!tudies using correlation
[ Positive correlation: When both variables increase in a study
[ Negative correlation: When one variable increases and the other
decreases
[ Co-efficients in correlation: A figure (number) demonstrating the
degree to which both variables are related (positive correlation
+1/negative correlation -1). The closer the correlation to a perfect
relationship (+1 or -1), the stronger the association between
variables
[ !tating correlations:
1. Zero (0 or closer to zero)
2. Weak (±0.1-±0.3)
3. åoderate (±0.4-±0.6)
4. !trong (±0.7-±0.9)
5. Perfect (+1/о1)
Example of a correlational analysis
ß     
 
1. Try to criticise this research, and think of the
arguments you might give to support the
claim that smaller class sizes are actually of
greater benefit to schoolchildren
2. Consider any extraneous or confounding
variables that might have affected the above
claim
Advantages of correlational designs
- Predictions more easily plausible: rf there has
been found a correlation, then we can make
predictions regarding the effect of one variable to
the other
- Allows quantification of relationships:
Correlations show the strength of a relationship
between two variables
- åanipulation is not possible: No requirement of
manipulating the behaviour of participants, so
correlations are ethically-prone to collecting data
and analysing them
Disadvantages of correlational designs
- Quantification problems: Although, correlations seem to appear quite low,
they can be meaningful or significant if the number of scores recorded is
quite high. Contrary to that, we may have a large figure in correlation, and
the results not to be significantly meaningful
- Cause and effect: We do not infer causation from correlation. Effects are
practically inexistent since the relationship between variables cannot
justify (be sure of) the direction of the association, for no r or D can be
selected
- Extraneous associations: There may be other variables, except the ones
studied; i.e. blowing wind and people͛s character traits in a normal winter
day, however what about summer temperatures?
- Only works for linear relationships: !traight-line relationships are tested
whether in a negative or positive correlation. rf we have a curvilinear
correlation it means we have aero significance in our results (correlation
coefficient/Pearson͛s r)
An example for discussion
[ Possible correlations:
1. Lower-income parents have more children
2. Older people make poorer eyewitnesses
3. !ales of running shoes have increased at the same time as
sales of personal computers
A. Think up of possible extraneous variables for the above
assumed correlations
°. Consider hypothetical correlations that might account for
a link between those variables
C. °y changing one variable with a confounding variable
(modifying our hypothesis) can we work out a more viable
correlation?
Observational techniques
1. Participant observation: °y becoming, the
experimenter a ͚participant͛ of his/her
participants͛ observation actual response to
the study
2. Non-participant observation: °y being the
researcher involved in the observation of
his/her participants͛ behaviour from a
distance
Advantages of observational
techniques
- High external validity: Observation takes place in
the natural environment; participants tend to
behave naturally, and results can be easily
generalised to other settings
- Practical method: !ocial context instead of
cooperation with participants: natural
environments where deliberate manipulations
would be unethical or impractical
- Fewer demand characteristics: Participants do
not know they are observed, so with their
responses they cannot satisfy the experimenter
Disadvantages of observational
techniques
- Cause and effect: With observations we do not have manipulations. Little
control over extraneous variables
- Observer bias: The question about reliable observation because of the
danger the results coming out of observations to be fabricated, it is cross-
checked with another͛s experimenter procedure of the same study
(inter/between reliability). Also, by checking all stages of the study and by
comparing them with one another from inception to completion, we look
for intra/within reliability
- Replication: Lack of proper control participants (proper, in terms of
repeating the same study again) leads to a stable problem of reliability and
validity of all observations (all of us do not possess the same ͚senses͛, or
͚sensitivity abilities͛ when observing situations)
- Ethics: rssues of informed consent and invasion of privacy arise when
participants are unaware of being observed
- Practical problems: rt is difficult for experimenters to remain unobserved,
so to avoid demand characteristics, or to categorise the observed
behaviours accurately
Practical learning: Classroom
observation
[ You are a government inspector of psychology classes.
You decide to use a non-participant observation
technique:
1. Outline exactly how you observe a psychology lesson
2. Consider the number of times students contribute to
the lesson; how frequently they take notes, their level
of engagement (behavioural categories). How would
you record the categories?
3. Observe a psychology lesson and write up your
experiences
4. Was the lesson accurately taught to students?
Questionnaires
1. Closed (fixed) questions or statements:
Usually ask for ticking boxes. Easy for
quantification, but are restriction-prone
2. Open questions/statements: Allowing
participants to answer in their own words.
They are difficult to analyse, but allow
freedom of expression
Advantages of questionnaires

[ Quick and clear


[ Large samples
[ Quantitative and qualitative analysis
[ Replication ʹ by the use of standardised
questions
!hortcomings of questionnaires
1. åisunderstandings: åisinterpretation of questions by
the participants
2. °iased samples: Needs time to be spent; asked from
people who are literate, or samples are chosen to
satisfy the findings expectation of the experimenter
(his/her hypothesis to be supported)
3. Low response rate: Not all questionnaires are
returned back
4. !uperficial issues: !ensitive issues arise that need
detailed understanding
5. !ocial desirability: Answers being presented in a
socially positive light
rnterviews
1. !tructured (formal) interviews: rdentical
questions addressed to participants in the
presence of the experimenter
2. Unstructured or informal interviews: Less
controlled and involve an informal discussion
as a particular topic. Questions are posed in
relation to participants͛ answers to the
subject
Advantages of interviews
[ Complex issues: Complicated or sensitive issues
are more easily to be dealt with in a face-to-face
interview
[ Ease misunderstandings: rnterviewers have the
chance to clarify any question of participants
during interview
[ Data analysis: °oth quantitatively and
qualitatively: ariety and flexibility to the
collection of data
[ Replication: The more structured an interview the
more easier to replicate
Disadvantages of interviews
[ rnterviewer or interviewee effects: °iased questions so
to elicit ͚fabricated͛ answers. rnterviewers may bias the
respondents͛ answers, but interviewees may develop
demand characteristics, and social desirability issues
[ rnterview training: structured interviews do not require
formal training. However, skills are needed for an
unstructured interview to be conducted
[ Ethical issues: rf participants do not know the true
purpose of the interview, ethical issues such as
deception and respect to privacy may arise
Think of a topic both of a quantitative
and qualitative research

[ Examples:
1. rnvestigating people͛s views on euthanasia
2. Questioning people͛s exposure to stressful
life events, and how they have experienced
their impact
!ummary No. 3
[ åost common aspects that arise in both
quantitative and qualitative research are:
1. Experimenter and Participant bias
2. !ocial desirability issues
3. !tructured and unstructured interviews
4. Participant observation and demand
characteristics
Reading a story and preparing it for
research
[ A research project on memory aimed to find
out about individuals͛ earliest memories. rn
order to do this the research team conducted
unstructured interviews with people aged 10
to 50 years. The interviewers started with a
set of standard questions, but then adapted
their questions in response to the answers
given by the interviewees
—etting ͚committed͛ as teachers in
research methodology
[ Research methodology is difficult also for us
teachers...
[ Assuming hypothetical studies while in class
with the students
[ Outlining the combination between theory
and research practice (  as a tool,
 as the means towards maintaining
findings)
TEȿɃɇ
(The End)

You might also like