Research Methods Revision

Rate your confidence on Research

Methods
• Experimental Method
• Experimental Design
• Types of Experiment
• Sampling
• Ethical Issues
• Correlations
• Observation
• Self-Report Techniques
• Types of Data
• Descriptive Statistics
• Graphs
• Statistical Testing
• Peer Review and the Economy
 Experimental Method
Key Concepts

Aim: A general expression of what the researcher intends to investigate

Hypothesis: A predictive statement of what the researcher believes they will find. It can be directional
(one-tailed) which states whether changes will be greater or lesser or non-directional (two-tailed) which
just predicts a difference/correlation

Variables Research Issues Research Techniques

Independent Variable: Demand Characteristics: Randomisation: The use of chance

the thing you change Any cue from the research when designing investigations to
to see the effect on situation or researcher control for the effects of bias
the DV that may reveal the aim of Standardisation: Using exactly the
Dependent Variable: the study and participants same formalised procedure for all
the thing you are act accordingly participants in a research study
measuring Investigator Effects: Control Groups: Used for the purpose
Extraneous Variables: Where investigators of setting a comparison. They act as a
other things that behaviour may affect the ‘baseline’ to help establish causation.
could affect the DV outcome of the research. Single Blind: Participant doesn’t
that you are not know the aims of the study
measuring Pilot Study Double Blind: Participant and
Confounding Variable: researcher doesn’t know the aim of
other variables that Trial run of the research to the study
have affected your
work out any problems
results
 Experimental Design
Independent Groups
Matched Pairs
One group do condition A
and a second group do Two groups of participants
Condition B. are used but they are
Participants should be related to each other by
Repeated Measures:
randomly allocated to being paired on participant
groups. variables that matter to
Same participants take part
the experiment.
in all conditions of the
EVALUATION: experiment.
EVALUATION:
The order of conditions
+ No order effects should be counterbalanced
+ Less likely to guess the + Participant Variables
to avoid order effects
aim + No order effects
- Participants Variables - Matching is not perfect
EVALUATION
- More participants used - More participants
+ Participant Variables
+ Fewer Participants
- Order Effects
- Participants may guess the
aim
 Types of Experiment
Laboratory Experiment Field Experiment
A controlled environment where extraneous and A natural setting. The researcher goes to the
confounding variables can be regulated. participants. The IV is manipulated and the effect on
Participants go to researcher. The IV is the DV is recorded.
manipulated and the effect on the DV is recorded.
EVALUATION
EVALUATION
+ More natural environment
+ EVs and CVs are controlled + Greater external validity
+ Can be easily replicated - More difficult to control CVs
- May lack generalisability - There may be ethical issues
- Demand Characteristics could be a problem
Quasi Experiment

Natural Experiment IV is based on a pre-existing difference between

people (age or gender). No one has manipulated
The experimenter does not manipulate the IV. The IV the variable, it simply exists. DV may be naturally
would have varied even if the experimenter wasn’t occurring or may be measured by the
interested. DV may be naturally occurring. experimenter

EVALUATION EVALUATION

+ Ethical + High control

+ Greater External Validity + Comparisons can be made between people
- Natural event may only occur rarely - Participants are not randomly allocated
- Participants are not randomly allocated - Causal relationships not demonstrated
 Sampling
Difference between Population and Sample

Population: A large group of people that a researcher is interested in investigating.

Sample: Usually not possible to include all members of the population in the study, so a smaller group is selected – the
sample

Opportunity Sample Random Sample

Volunteer Sample
Every person in the target population has an
People who are the most
equal chance of being selected. E.g. pulling names
available and willing at the Participants select themselves out of a hat
time for the study
Evaluation
EVALUATION: EVALUATION
+ Unbiased
+ Quick + Participants are willing - Not necessarily representative
- Biased - Likely to be biased

Stratified Sample
Systematic Sample
Participants are selected according to their
Participants are selected using a set ‘pattern’. frequency in the target population. E.G a strata (sub-
E.g. every nth person is selected from a list groups) is identified and then a random sample is
of the target population selected from each

EVALUATION EVALUATION

+Unbiased + Representative
- Time and effort - Strata won’t reflect all personal differences
Informed Consent Ethical Issues Protection from Harm

Participants should be Participants should be at no

able to make an informed more risk than they would be
judgement about whether Deception in everyday life.
to take part.
Deliberately misleading or They should be given the
Too much information withholding information so right to withdraw at any
may affect participants consent is not informed. time, they should be
behaviour so alternative reassured that their
forms of consent are: ask Participants should be debriefed behaviour was typical/
a similar group, they to be told the true aim of the normal when being
agree to be deceived, or study, details that were not given, debriefed and the researcher
get consent after the study what their data will be used for should provide counselling if
and their right to withhold the participants have been
information distressed

Privacy/Confidentiality

Participants have the right to control information about themselves, if that is invaded confidentiality
should be rightfully respected.

If personal details are held they must be protected. Usually no personal details are recorded.
Researchers refer to participants using numbers, initials or false names.
Participants personal data cannot be shared with other researchers.
 Correlations
Differences between Correlation and an Experiment

Experiment: Researcher manipulated the IV and records the effect on a DV.

Correlation: No manipulation of variables and so cause and effect cannot be demonstrated.
The influence of EVs is not controlled.

Association: Types of Correlation:

Correlations illustrate the Positive: as one variable

strength and direction of increases, the other
an association between increases
two co-variables. Negative: as one
variable increases, the
They are plotted on a other decreases
scattergram Zero: No relationship
between the two EVALUATION
variables
+ Useful starting point for research
+ Relatively economical
- No cause and effect
- Method used to measure
variables may be flawed
Naturalistic / Controlled

Naturalistic: Takes place where the

Observation Participant / Non-Participant

Participant: Researcher becomes a

target behaviour would normally occur Covert / Overt part of the group they are studying

EVALUATION Covert: Participants are unaware EVALUATION:

+ High external validity they are being studied. + Lead to greater insight
- Low control - Loss of objectivity
EVALUATION
Controlled: Some control/manipulation + Demand characteristics reduced Non-Participant: Researcher
of variables including control of EVs - Ethically questionable remains separate from the group

EVALUATION Overt: Participants are aware of EVALUATION

+ Can be replicated being studied + More objective
- May have low external validity - Loss of insight
EVALUATION:
+ More ethically acceptable
- Demand Characteristics

Observational Design:
Behavioural Categories Time Sampling
Event Sampling
The target behaviour to be observed Observations are made at regular
should be broken up into a set of intervals, e.g. once every 15 A target behaviour/event is recorded
observable categories. This is similar seconds every time is occurs
to the idea of operationalization
EVALUATION: EVALUATION
EVALUATION + Reduced the number of + May record infrequent behaviour
- Difficult to make clear observations - Complex behaviour is oversimplified
- Dumped behaviours go unrecorded - May be unrepresentative
Questionnaires
Self-Report Techniques
Interviews
Made up of a pre-set list of written questions (or
items) to which a participant responds. They can be
Face to face interaction between interviewer and interviewee
used as part of an experiment to assess the DV.
Structured Interview: list of pre-determined questions asked
Evaluation:
in a fixed order
+ Can be distributed to lots of people
EVALUATION:
+ Respondents may be willing to ‘open up’
+ Easy to replicate
- Responses may not always be truthful
- Interviewees cannot elaborate
- Response bias
Unstructured Interview: There are no set questions. There is
Designing Questionnaires
a general topic to be discussed but the interaction is free-
Write good questions: avoid jargon, avoid double-
flowing and the interview is encouraged to elaborate
barrelled questions and avoid leading questions.
EVALUATION:
Closed Questions: Respondent has limited choices.
+ There is greater flexibility
- Difficult to replicate
EVALUATION:
+ Easier to analyse
Semi-Structured Interviews: List of questions that have been
- Respondents are restricted
worked out in advance but interviewers are free to ask
follow-up questions when appropriate.
Open questions: Respondents provide their own
answers expressed in words
Designing Interviews
Have an interview schedule, a quiet room, rapport with the
EVALUATION:
participants and make sure that you are abiding by ethics.
+ Respondents not restricted
- Difficult to analyse
 Types of Data
Quantitative / Qualitative Primary Data / Secondary Data / Meta-Analysis

Quantitative Data: Numerical data. E.g. Primary Data: ‘first hand’ data collected for the
purpose of the investigation
reaction time or number
EVALUATION
EVALUATION: + Fits the job
+ Easier to analyse - Required time and effort
- Oversimplifies behaviour
Secondary Data: Collected by someone other than
Qualitative Data: Non-numerical data the person who is conducting the study.
expressed in words e.g. extract from a
diary EVALUATION
+ Inexpensive
- Quality may be poor
EVALUATION
+ Represents complexities Meta-Analysis: A type of secondary data that
- Less easier to analyse involves combining data from a large number of
studies

EVALUATION
+ Increases validity of conclusions
- Publication Bias
 Descriptive Statistics
Measures of Central Tendency Measures of Dispersion
Mean: arithmetic average, add up all the
Range: The difference between the highest
scores and divide by the number of scores
and lowest value (+1)
EVALUATION
+ Sensitive EVALUATION:
- May be unrepresentative + Easy to calculate
- Does not accoumt for the distribution of
Median: Middle value, place scores in scores
ascending order and select middle value. If
there are two middle values, the mean is Standard Deviation: Measure of the average
calculated.
spread around the mean. The larger the
EVALUATION standard deviation, the more spread out the
+ Unaffected by extreme scores data is.
- Less sensitive than the mean
EVALUATION:
Mode: Most frequent or common value. + More precise than the range
- It may be misleading
EVALUATION:
+ Relevant to categorical data
- An overly simple measure
Presentation of Quantitative Data Graphs
Tables: Raw scores are displayed in
columns and rows; a summary paragraph Distributions:
beneath the table explains the results
Normal Distribution: Symmetrical, bell-
Bar Chart: Categories are usually placed shaped curve. Most people are in the middle
along the x axis and frequency on the y area of the curve with very few at the extreme
axis. The height of each column ends. The mean, mode and median all occupy
represents the frequency for that item. the same mid-point of the curve.

Histogram: Bars touch each other – data Skewed Distribution: Distributions that lean to
is continuous rather than discrete. There one side or the other because most people
is a true zero. are either at the lower or upper end of the
distribution.
Line Graph: Frequency on one axis, data Positive Skew: Most of the distribution is
on the other axis is continuous. The line concentrated towards the left of the graph,
often shows how something changes resulting in a long tail on the right.
over time Negative Skew: Most of the distribution is
concentrated towards the right of the graph,
Scattergram: Used for correlational resulting in a long tail on the left.
analysis. Each dot represents one pair of
related data. The data on both axis must
be continuous.
 Statistical Testing
Statistical Testing: To Find the Critical Value You Need to Know:
The significance level (0.05)
The number of participants in the investigation (N value)
Significance: The difference/
Whether the hypothesis is directional or non-directional
association between two sets of data
is greater than what would occur by The Sign Test
chance. To find out if it is significant
we need to do a statistical test.
Used to analyse the difference in scores between
Probability: A numerical measure of
related items (same participants used twice)
the likelihood that cvertain events will
occur. The accepted level of Calculation:
probability in psychology is p≤0.05 The score for condition B is subtracted from
(there is a 5% chance that the results Condition A to produce a sign of difference (+ or -).
are down to coincidence) Do this for each participant
The total number of pluses and the total number of
Calcuated Value: the value you minuses should be calculated
calculate through your statistical test
Participants who achieved the same score in both
and compare to a critical value to
decide whether it is significant or not. conditions should be disregarded and deducted from
the N value
Critical Value: Given in a table of The S value is the total of the less frequent sign
critical values.
Critical Value: If the critical value is equal to or less
than you S value then it is significant.
 Peer Review and the Economy
Peer Review Psychology and the Economy
Before publication, all aspects of the The findings of psychological research can benefit
investigation are scrutinised by experts our economic prosperity.
(peers) in the field. These experts should
be objective and unknown to the Attachment Research into the Role of the Father
researcher Recent research has stressed the importance of
multiple attachments and the role of the father in
healthy psychological development. This may
The Aim: To allocate research funding;
promote more flexible working arrangements in the
validation of the quality and relevance of family and means that modern parents are better
research and improvements and equipped to contribute more effectively to the
amendments are suggested. economy

EVALUATION: Development of Treatment for Mental Illness

+ Protects quality of published research A third of all days off work are caused by mental
- May be used to criticise rival research illness. Psychological research into the causes and
- Publication bias treatments means that patients can have their
disorder diagnosed quickly. Patients have access to
- Ground-breaking research may be buried
therapies or psychotherapeutic drugs and sufferers
can manage their condition effectively, return to
work and contribute to the economy.