Research Methods Flashcards
1
Q
Independent variable (IV)
A
- In experiments
- Manipulated by the experimenter in order to observe effects on dependent variable
- E.g. what are the two conditions of the experiment? - male or female
- This is used to establish cause and effect
- The one that is CHANGED
2
Q
Dependent variable (DV)
A
- In experiments
- The DV depends in some way on the IV
- How the effect of the change in the IV is measured
- E.g. different scores, different times
- The one that is MEASURED
3
Q
Co-variables
A
- In correlations
- When using a correlation, it is impossible to directly manipulate the variables
- In this circumstance we would have to measure two separate variables and then compare them to each other
4
Q
Operationalisation
A
Making something measurable
5
Q
Examples of operationalisation of variables
A
1 - stating the method of measuring the DV
DV is time = not operationalised
DV is time takes to do task in seconds = operationalised
2 - Using a questionnaire to measure happiness
6
Q
Extraneous variables
A
Any variables that might affect the DV so we try to keep them the same to ensure a fair test
7
Q
Confounding variables
A
Variables which have not been kept the same (controlled) so may have affected the DV
8
Q
Experiments
A
Involve manipulation of an independent variable to measure the effect on a dependent variable. There are 4 types of experiment.
9
Q
Lab experiment
A
- IV is directly controlled by the experimenter
- Takes place in tightly controlled, artificial situations
10
Q
Strengths of lab experiments
A
- Easy to replicate
- High internal validity due to good control of variables
- Can establish cause and effect
11
Q
Limitations of lab experiments
A
- Lacks ecological validity as tasks are more artificial
- Participants may try to guess the aim of the study
12
Q
An example of a lab experiment
A
Lotus and Palmer
13
Q
Field experiment
A
- IV is deliberately manipulated by the researcher
- A controlled experiment
- It is conducted in a more “ordinary” environment
- The “field” is anywhere outside of the laboratory
14
Q
Strengths of field experiments
A
- Participants are usually not aware they are participating in an experiment so behaviour may be more natural as they are not responding to demand characteristics
- Participants may be more relaxed
- Higher external validity due to greater mundane realism
15
Q
Limitations of field experiments
A
- Lower internal validity (more difficult to control extraneous and confounding variables)
- Ethical issues as participants may not know they are being studied
16
Q
Natural experiment
A
- Conducted when it is not possible, for ethical or practical reasons, to deliberately manipulate an IV
- The IV varies naturally and would vary whether or not the researcher was interested
17
Q
Strength of natural experiments
A
- Enables psychologists to study “real” problems such as the effects of a disaster on health (mundane realism and ecological validity)
18
Q
Limitations of natural experiments
A
- Cannot demonstrate causal relationships because IV is not directly manipulated
- Random allocation is not possible so confounding variables may impact internal validity
- Can only be used where conditions vary naturally
19
Q
Quasi experiment
A
- Studies that are “almost” experiments
- IV is actually not something that varies at all - it is just a condition that exists e.g. age and gender
20
Q
Strength of quasi experiments
A
- Allows comparisons between types of people
21
Q
Limitations of quasi experiments
A
- Participants may be aware of being studied, creating demand characteristics and reducing internal validity
- Dependent variable may be a fairly artificial task, reducing mundane realism
22
Q
Example of a field experiment
A
Bickman (IV - what they were wearing (dressed as a milkman, security guard and ordinary clothes) DV - obedience)
23
Q
Example of a natural experiment
A
Effects of institutionalisation in Romanian orphanages
Rutter (IV - age of adoption, DV - IQ)
24
Q
Example of a quasi experiment
A
Miller (1956)
How age affects short-term memory (digit span test)
25
What chart is used for experiments?
Bar chart
26
Which graph is used for correlations?
Scattergraph
27
What do correlations not look at?
Whether an IV affects a DV
28
Correlations
Allow a psychologist to examine the relationship between two co-variables. These co-variables exist and are not manipulated by the psychologist.
29
Correlation coefficient
Tells us whether a relationship is positive or negative, and tells us how strong the relationship is. It is a number between -1 and +1.
30
Strengths of correlations
- Psychologists can study a topic where it would be unethical to carry out an experiment
- The strength of a relationship can be found by using scattergraphs and calculating the correlation coefficient
31
Limitations of correlations
- Can’t establish cause and effect between the co-variables i.e. we can’t say that changing one variable causes the other variable to change
- Another variable may cause both variables to change
32
Observational study
A study that involves observing actual behaviours
33
Controlled observation
The setting for the observation is structured and controlled e.g. Ainsworth’s strange situation
34
Naturalistic observations
- Setting for an observation is natural
- Designed to examine behaviour without the experimenter interfering with it
35
What might a naturalistic setting do to demand characteristics?
Reduce them as people will act more naturally
36
Strength of naturalistic observations (will be the opposite for controlled observations)
- People tend to behave more naturally than for controlled observations (providing higher external validity)
- The information gathered tends to be richer and fuller than experimental methods
37
Limitations of naturalistic observations (will be the opposite for controlled observations)
- Researcher has no control over the situation. Low internal validity and reliability, compared to controlled observations
- Difficult to replicate
38
Covert observations
Participants’ behaviour is watched and recorded without their knowledge and consent
39
Overt observations
Participants’ behaviour is watched and recorded with their knowledge and consent
40
Strength of covert observations (will be the opposite for overt observations)
Should be higher in validity as people may change their behaviour if they are being observed
41
Limitation of covert observations (will be the opposite for overt observations)
Raise significantly more ethical issues regarding privacy as you shouldn’t observe covertly where people would not expect to be seen
42
Hawthorne effect
People change their behaviour when they know they are being observed
43
Participant observations
The researcher observes the participants from within the group that they are observing
44
Non-participant observations
The researcher observes the group of participants from a location away from the group
45
Strength of participant observations (will be the opposite for non-participant observations)
They give a greater insight into the behaviour of the group
46
Limitation of participant observations (will be the opposite for non-participant observations)
They are less likely to be objective
47
Why might researchers lose objectivity in a participant observation?
Researchers may form relationships with those they are observing
48
Unstructured observations
The researcher records all relevant behaviour but has no system
49
Problems with unstructured observations
- There may be too much to record
- The behaviours recorded will often be those which are most visible or eye-catching to the observer but these may not be the most important or relevant behaviours
50
A situation where an unstructured observation would be used
Where research has not been conducted before as a kind of pilot study to see what behaviours might be recorded using a structured system
51
Structured observations
- Aim to be objective and rigorous
- A researcher uses various systems to organise observations such as behavioural categories and sampling procedures
- More common in psychology
52
Behavioural categories
Dividing a target behaviour (e.g. stress) into a subset of specific and operationalised behaviours
53
Behavioural categories should…
- Be objective (no inferences need to be made about behaviours)
- Cover all possible component behaviours and avoid a “waste basket” category
- Be mutually exclusive (you should not have to mark two categories at one time)
54
Event sampling
Counting the number of times a certain behaviour (event) occurs in a target individual or individuals e.g. counting how many times a person smiles in a 10 minute time period
55
Time sampling
Recording behaviours in a given time frame e.g. noting what a target individual is doing every 30 seconds or some other time interval. At that time the observer may tick one or more categories from a checklist
56
Interviews
- A face-to-face interaction that results in the collection of data
- Questions can be predetermined or created in response to answers
57
Strengths of interviews
- A lot of data can be collected
- Face-to-face can assess emotions
58
Limitations of interviews
- Social desirability bias and interview bias
- Requires skilled personnel
59
Social desirability bias
People want to look good so they may be dishonest
60
Interviewer bias
The way the interviewer treats the participants might affect their answers
A biased interviewer may ask leading questions or otherwise affect the behaviour of participants, reducing validity
61
Structured interview
Pre-determined questions i.e. a questionnaire that is delivered face-to-face with no deviation from the original questions
62
Unstructured interview
New questions are developed as you go along
63
Strength of a structured interview
Easier to compare people’s answers
64
Strength of an unstructured interview
More natural so people may reveal more
65
Recording the interview
- Taking notes may interfered with the interviewer’s listening skills
- Taking notes may also make the respondent feel like they’re being evaluated especially if the interviewer doesn’t write everything down
- Best to record the interview and transcribe later
66
What effect can having an interviewer who is interested in the respondent’s answers have?
It may increase the amount or the quality of the information
67
Questionnaire survey
- Data that is collected using a set of written questions
68
Strengths of questionnaires
- A lot of data collected
- Does not require specialist administrators
- More anonymous
69
Limitations of questionnaires
- Leading questions and social desirability bias
- Biased samples
70
Clarity (questionnaires)
The respondent needs to know exactly what is being asked. There should be no ambiguity.
71
Bias (questionnaires)
The questions should not lead the respondents to a particular answer e.g. social desirability bias
72
Analysis (questionnaires)
The answers need to be easy to analyse
73
Open questions
Harder to analyse due to the wide range of answers
74
Closed questions
Have a limited range of answers e.g. yes or no, scales etc. and are therefore easy to analyse
75
Filler questions (questionnaires)
Including some irrelevant questions helps distract respondents from the true purpose of the study
Prevent demand characteristics
76
Sequence for the questions (questionnaires)
It is better to start with easier questions as it puts respondents at ease
77
Sampling technique (questionnaires)
Questionnaires often use stratified sampling, which makes the answers more representative
78
Pilot study (questionnaires)
Test the questionnaires on a small group of people so they can be refined later
79
Case studies
An in depth study that gathers a lot of detail about one person or a small group
80
Examples of case studies
- Phineas Gage
- HM
- Tan
81
Why is data triangulated?
To form a consistent conclusion about the case
82
Strengths of case studies
- Good ecological validity (real life application)
- A more ethical way of studying psychology (can’t inflict pain or trauma so take the opportunity)
83
Limitations of case studies
- Cannot generalise
- Cannot replicate
- Time consuming/ have to employ people for an extended period of time
84
Content analysis
- A technique for systematically summarising and describing any form of content-written, spoken or visually
- It converts qualitative data to quantitative data
85
5 steps for content analysis
- Data is collected
- Researcher reads through or examines the data, making themselves familiar with it
- The researcher identifies coding units
- The data is analysed by applying the coding units
- A tally is made of the number of times that a coding unit appears
86
Thematic analysis
- The researcher closely examines the data to identify common themes, topics, ideas and patterns of meaning that come up repeatedly
- Psychologists may select quotes from the material they are analysing to exemplify the themes they’ve identified
87
Strengths of content analysis
- Reliable way to analyse qualitative data as the coding units are not open to interpretation and so are applied in the same way over time and with different researchers
- It is an easy technique to use and is not too time time consuming
- It allows a statistical analysis to be conducted if required as there is usually quantitative data as a result of the procedure
88
Limitations of content analysis
- Causality cannot be established as it merely describes the data
- As it only describes the data it cannot extract any deeper meaning or explanation for the data patterns arising from
89
Aim
A statement of what you want to find out e.g. are males or females better drivers?
90
Hypothesis
A formal prediction of what you think will happen. It must include the operationalised IV and DV e.g. male drivers will have fewer faults on a driving test than female drivers
91
A directional hypothesis/one tailed
Predicts the kind of difference between two conditions in an experiment or the direction of the correlation
E.g There will be a positive correlation between number of homeworks completed and final grade
92
When do we use a directional hypothesis?
When there is previous research to support the prediction
93
A non-directional hypothesis
Simply predicts that there will be a difference or a correlation but doesn’t state what the direction will be e.g. there will be a difference between number of error made by males and females in a driving test
94
When do we use a non-directional hypothesis?
When there is no previous research or when there are mixed findings
95
Target population
A group of people that are the focus of the researcher’s interest
96
Sample
The group who take part in the research. We want the sample to be representative of the target population
97
Random samples
Everyone in the entire target population has an equal chance of being selected
Put names in a hat and draw out without looking
98
Strengths of random sampling
- It is widely accepted that since each member has the same probability of being selected, there is a reasonable chance of achieving a representative sample
- It is an unbiased method
99
Limitations of random sampling
- It can be impractical (or not possible) to use a completely random technique e.g. the target group may be too large to assign numbers to
- The people you select may not want to take part
100
Opportunity sampling
Uses people from target population available at the time a willing to take part. It is based on convenience.
101
Strength of opportunity sampling
It is easy and inexpensive to carry out
102
Limitation of opportunity sampling
The consequent sample may not be representative as it could be subject to bias - you are only likely to approach certain types of people to ask if they want to take part in
103
Systematic sampling
Chooses subjects in a system (i.e. orderly/logical) way from the target population e.g. every nth participant on a list of names
104
Strength of systematic sampling
Assuming the list order has been randomised, this method offers an unbiased chance of gaining a representative sample
105
Limitation of systematic sampling
If the list has been assembled in any other way, bias may be present e.g. if every 4th person in the list was male, you would only have males in your sample
106
Volunteer sampling
Individuals choose to be a part of the study e.g. they respond to an advert
107
Strength of volunteer sampling
- Often achieves a large sample size through reaching a wide audience e.g. with online advertisements
- The people want to take part
108
Limitations of volunteer sampling
Those who respond to the call for volunteers may all display similar characteristics (such as being more trusting or cooperative than those who did not apply) thus increasing the chances of yielding an unrepresentative sample
109
Stratified sampling
The researcher identifies the different types of people that make up the target population and works out the proportions needed for the sample to be representative
110
Strength of stratified sampling
It is the only way to get a true representative sample
111
Limitations of stratified sampling
- It takes more time and resources to plan
- Care must be taken to ensure each key characteristic present in the population is selected across strata, otherwise this will design a biased sample
112
Experimental design
Refers to how psychologists use participants within an experiment
113
Repeated measures
Involves using the same subjects in each condition of an experiment e.g. giving a group of subjects no alcohol followed at a later time by the same test after a pint of lager
114
Strengths of repeated measures
- No participant variables (e.g. both groups have the same driving ability as they are the same people)
- Less participants are needed
115
Limitations of repeated measures
- Order/practice effects (participant may get better second time due to practice)
- Demand characteristics
116
Counterbalancing
An attempt to control order effects in a repeated measures design. Half the participants take part in A then B then the other half do B then A
117
Independent measures
Involves using DIFFERENT subjects in each condition of the experiment e.g. giving one group of subjects a driving test with no alcohol and a different group of subjects the same test agates a pint of lager
118
Strengths of independent measures
- No order effects
- Less demand characteristics
119
Limitations of independent measures
- Participant variables
- More participants needed
120
Matched pairs
Involves using different but similar subjects in each condition of experiment. An effort is made to match the subjects in each condition in any important characteristics that might affect performance e.g. in driving ability, alcohol tolerance etc.
121
Strengths of matched pairs
- No order effects
- Less demand characteristics
- Less participant variables
122
Limitations of matched pairs
- Time consuming
- Difficult to do well
123
Order effects
Order effects can occur in a repeated measures design and refers to how the positioning of tasks influences the outcome e.g. practice effect or boredom effect on second task
124
Non-experimental design
How we design observations, interviews and questionnaires
125
Consistency
The same people doing the same thing should get the same result
126
A reliable observation should have…
A positive correlation between 2 observers watching the same thing at the same time
127
Pilot studies
- A small scale trial to help forsee any problems with a study
- Allow us to check that participants know what they have to do and that the procedure is fit for purpose
- Researchers can ask participants of pilot studies for feedback
- Potentially saves time and money in the long run that could’ve been wasted on a badly designed study
128
Demand characteristics
Where participants guess the aim and act up to the situation. Participants behave how they think you want them to in order to be a ‘good participant’
129
Screw you effect
Participants acting in a way that researchers do not want on purpose
130
Solution for demand characteristics
Deception/single blind technique
131
Investigator effects
How the researcher’s behaviour might affect the participants. They could show bias as they want to see the result they expect
132
Solution for investigator effects
Double blind technique where a 3rd party conducts the research who doesn’t know the aims.
133
Ethical issue
These arise when there is a conflict between the rights of the participant and the goal to produce authentic results
134
Ethical guidelines
Produced by the BPS to help researchers deal with the dilemmas.
135
Informed consent
Making participants aware of the aims, procedure and their rights before the study goes ahead
136
Ways of dealing with informed consent
- All participants should sign a consent form
- Parental consent for those under 16
Not ideal methods:
- Presumptive consent (ask a similar group of people)
- Prior general consent (give consent to a number of studies)
- Retrospective consent (get consent after the study)
137
Deception
Deliberately misleading or withholding information from participants. This makes informed consent impossible.
138
Ways of dealing with deception
Debrief participants at the end (inform them what happened, what their data will be used for and allow opportunity to withdraw results)
139
Protection from harm
Participants should not be placed at any more risk than normal daily life - this includes stress and embarrassment
140
Ways of dealing with protection from harm
- Reassure participants that their behaviour was normal
- Offer counselling if effects are extreme
- Follow up
141
Privacy and confidentiality
This is in line with data protection laws which is the right to have personal data protected
142
Ways of dealing with privacy and confidentiality
- Maintain anonymity - don’t record names - use numbers
- Don’t share data with other researchers
143
Ethics committee
Group of psychologists who weigh up the costs and benefits of a research proposal
144
Consent forms
- Participants need to sign a consent form to take part in a study
- Set out as a formal letter
- Includes procedural details and ethical rights
145
Scientific method
Observe - develop a theory - create a precise testable hypothesis - research - accept or reject hypothesis - change or support theory
146
Replicable
Easily repeatable
147
Objective
Free from bias/emotion
148
Controlled
Fair testing
149
Title
Concise, precise and to the point
150
Abstract
- 100-150 word summary of aim, method, results and conclusion
- Quick to read and understand
151
Introduction
- Often the longest section
- Literature review of previous research in that area
- States aim and hypothesis at the end of
152
Method
- Should be in enough detail to allow someone to replicate the research
- Outlines design, sample, materials and procedure
153
Results
- A summary of the data, section of descriptive statistics, analysis of results of texts
- Finishes with rejection or acceptance of null hypothesis
154
Discussion
- Discusses anything that went wrong or the limitations of the study
- Suggests how it could be improved if it were to be repeated
- Suggestions for alternative studies and future research are explored
- Ends with a paragraph summing up what was found and assessing the implications of the study and any conclusions that can be drawn from it
- Discusses how research compares to previous research
155
References
- Look through the report and include a reference for every researcher mentioned so people can find the original sources
- Should be in alphabetical order
156
Order for writing references
Name, date, title, book/journal published
157
Appendices
- Comes at the end of a report
- Contains any additional information e.g. raw data or interview transcripts
- Information is relevant but is too long or too detailed to include in the main body of your work
158
Peer review
The way that psychological research findings are shared with other scientists and are held up for scrutiny
159
Steps of peer review
1) A research paper is submitted to a journal is considered to be worthy of publication
2) The editor sends this paper to other experts in the field who critically appraise all aspects of the study then return it with their recommendations as to whether the work is of acceptable quality
3) If not, researchers revise their work and resubmit their paper
160
Limitations of peer review
- Peer review can act to maintain the status quo and prevent potentially revolutionary research from being published because science is conservative and resistant to large changes changes in opinion. If a study doesn’t fit with accepted existing knowledge it can be rejected
- A reviewer may strongly support an opposing view making them less likely to provide an unbiased opinion of the work
- Not necessarily anonymous as the research world is very small
- Institution and gender bias
- The ‘file-drawer’ problem - a bias towards publishing studies with positive results (supporting the hypothesis) For every positive finding there could be hundreds of negative results
161
Institution bias
Research from prestigious universities is favoured
162
Gender bias
Male researchers tend to be favoured
163
Qualitative
Describes qualities and characteristics
164
Examples of qualitative data
Interviews, unstructured observations
165
Weakness of qualitative data
Less easy to analyse
Hard to draw conclusions
166
Quantitative
Numerical
167
Example of quantitative data
Reaction time
Number of mistakes
168
Strengths of quantitative data
Objectivity
Clarity
Predicative capabilities
169
Weaknesses of quantitative data
- Oversimplify or ignore meanings
- Potential for researcher bias
- Limited ability to capture complex human experiences
170
Primary data
Collected first hand for the purpose of an investigation
171
An example of primary data
Milgram
172
Strengths of primary data
More accurate and reliable because it comes from a direct source
Can be collected in real time
Faster to collect
173
Weaknesses of primary data
Time and effort
174
Secondary data
Collected for a different purpose
Collected by someone other than the secondary user
175
An example of secondary data
Government publications
176
Strength of secondary data
Inexpensive
Less effort
177
Weaknesses of secondary data
Lack of relevance - rarely provides the answers needed
Lack of accuracy
178
Meta analysis
Examination of data from a number of independent studies of the same subject, in order to determine overall trends
179
An example of a meta analysis
Van Ijzendoorn and Kroonenberg (Cultural variation in attachment)
180
Strength of meta analysis
Increases validity of conclusions — much larger sample size
181
Weaknesses of meta analysis
Publication bias - not all studies will be selected
182
Descriptive statistics
Describe the data collected
183
Measures of central tendency
Averages (mode, median, mean)
184
Measures of dispersion
How spread out the data is (range, standard deviation)
185
How to find the median
Put in size order and find the middle value
186
Normal distribution curve
Mean, mode and median are the same
187
Negative skew
Mode is greater than the mean (left foot)
188
Positive skew
Mode is less than the mean (right foot)
189
The bigger the standard deviation…
The more spread out the data is
190
What percentage of people lie within 1 standard deviation of the mean?
68%
191
What percentage of people lie within 2 standard deviation of the mean?
95%
192
Inferential statistics
Tell us if the difference or correlation we have found in our research significant or if it is just a fluke/chance result
193
When do we accept the hypothesis? (Inferential statistics)
If the results are significant at the 5% level (i.e. there is a 5% or less chance they were a fluke)
194
Things to consider when using a critical value table
- Is the hypothesis directional or non-directional?
- Number of participants
- Look down the 5% (0.05) column
195
How do you know if the findings are significant?
If the critical value is equal to or less than 1
196
Sign test step by step
1) Cross out any no difference scores and discount these participants from n
2) Count the number of + and -
3) Find out whether the number of + or - are the lowest
4) The lowest value is the S value
5) Use the critical value table to find the critical value for this data set at the 5% level
6) If the S value is equal to or less than the critical value then the research is significant
197
Reliability
How consistent the data is
198
Inter-observer reliability
When another observer repeats the test and compares their results with yours to see if you have high agreement (1) or low (0) This is a kappa score. To improve this score you can include/amend your behaviour categories
199
Test-retest
Giving the same group of participants the same test at a different time and assessing the score similarity. This can be improved by making your test question detailed and specific
200
Standardisation
To ensure that each procedure is robust and repeated consistently across trials. This will improve reliability
201
Validity
How accurate the data is
202
Ecological validity
The ability to generalise the research results to different environments and achieve the same results
203
Mundane realism
How realistic are the tasks to the real world e.g. counting backwards in 3s
204
Temporal validity
The ability for the research results to be generalised to different time periods e.g. Asch
205
Population validity
Can the results be generalised to other samples of participants
206
Concurrent validity
To compare your research results to other similar results in the field and assessing if they’re similar findings
207
Face validity
To the extent in which the test measures what it claims to measure e.g. IQ test - intelligence or memory
208
5 features of a science
Empirical methods (lab studies)
Objectivity
Replicability
Theory construction
Hypothesis testing
209
Falisifiability
Always aiming to prove your hypothesis wrong
210
Paradigm
A set of ideas which can change over time
211
Paradigm shift
A sudden major shift in thinking that requires revolution in thinking
212
Nominal data
Data in named categories e.g. males, females, football teams
213
Ordinal data
Data that can be ordered via rank or a rating scale
214
Interval data
Data with equal measurements in-between each value e.g. time, temperature
215
Type 1 error
False positive. I’ve rejected the null hypothesis when I should’ve accepted it. You believe you have found a genuine positive effect when there isn’t one e.g. a male being pregnant because they have the symptoms
216
Type 2 error
You fail to reject the null hypothesis (you accept it) and believe there isn’t a negative effect when there is one. E.g. a pregnant female being told she’s not pregnant because of other factors
217
Null hypothesis
An assumption that there is no relationship/ difference/ association. When conducting research we aim to reject our null hypothesis (falsifiability)
218
What is reliability?
Refers to the consistency of a research study or measuring test
The same people doing the same thing to get the same results
219
Two key words when thinking about reliability
Same
Scattergraph
220
How is external reliability assessed?
Test-retest method or inter-rater reliability
221
How is internal reliability assessed?
The split half method
222
Split half method
In split-half reliability, the test or measurement instrument is divided into two halves or subsets of items, and the scores on each half are compared. If scores are similar then the test has good internal reliability.
223
What is the correlation coefficient from a test-retest scattergram that means there is good reliability?
+0.7 (more than 70% agreement)
224
How can we improve reliability?
- Operationalise behaviour categories so that observers can consistently identify things correctly
- Pilot studies can give observers insights as to what will work and what needs to be improved to ensure reliability
- Standardised instruction so that other researchers can repeat scientific procedures and findings
- Training in observation techniques being used and making sure everyone agrees and fully understands the processes involved
- Record data accurately by ensuring methods are easy to use
225
External validity
About generalisation - can we generalise the results beyond the study itself?
226
How to improve external validity
- Representative sample
- Realistic settings/tasks
227
Internal validity
About measurement - the ability of a study to test the hypothesis it was designed to test
228
Factors affecting internal validity
- Confounding variables
- Poor operationalisation
- Bias - demand characteristics - social desirability bias - investigator effects - confirmation bias
229
Confirmation bias
Researchers only see what they expect to see and ignore other behaviours
230
What kind of data is the mode used for?
Nominal data
231
What kind of data is the median used for?
Ordinal data
232
What kind of data is the mean used for?
Interval data
233
When should you not use the mean?
If there are extreme outliers
234
On a normal distribution curve, what three things are the same?
Mean
Mode
Median
235
Statistical tests
- Used to determine whether a difference or correlation found in an investigation is statistically significant
- The outcome of statistical tests allow us to either accept or reject the null hypothesis
236
What does the choice of test depend on?
The type of investigation: correlational or experimental
The experimental design: independent groups, repeated measures or matched pairs
The level of measurement (type of data): nominal, ordinal, interval
237
Nominal data, experiment, unrelated design
Chi-squared
238
Ordinal data, experiment, unrelated design
Mann-Whitney
239
Interval data, experiment, unrelated design
Unrelated t-test
240
Nominal data, experiment, related design
Sign test
241
Ordinal data, experiment, related design
Wilcoxon
242
Interval data, experiment, related design
Related t-test
243
Nominal data, correlation
Chi-squared
244
Ordinal data, correlation
Spearman's rho
245
Interval data, correlation
Pearson's r
246
Implications
How does what we learn from findings of psychological research influence, affect, benefit or devalue our economic prosperity
247
Examples of implications of psychological research for the economy
- Funding institutions like law enforcement, mental health, social work
- Reducing cost of treatment/maintenance for those institutions
248
Three key factors to consider when thinking about the economy
Taxes
Employment/productivity
NHS
249
The McCrone report (2008)
- Compared the uses of different forms of therapy
- Specifically drugs vs psychotherapies
- Drug treatments provide a much greater economic gain than other therapies, that might be as effective, but much more expensive to run
- Drug therapies are also quicker allowing people to return to work sooner
250
The McCrone report (2008)
- Compared the uses of different forms of therapy
- Specifically drugs vs psychotherapies
- Drug treatments provide a much greater economic gain than other therapies, that might be as effective, but much more expensive to run
- Drug therapies are also quicker allowing people to return to work sooner
251
The smaller the standard deviation…
The more clustered (less spread out/variable) the values are around the mean (scores are more similar)
252
How would you obtain the male participants for a stratified sample of 20 people?
There are 60 men and 40 women in the target population
- Put all 60 male names in a hat
- Determine the proportion of males needed to mirror the number of males in the target population (60% because the males are 60 of the 100 people in the target population)
- Calculate 60% of 20 = 12
- Draw out 12 names
253
The bigger the mean…
The more spread out the data
254
How does using standard deviation rather than the range improve a study
- Standard deviation is a measure of dispersion that is less easily distorted by a single extreme score
- It takes it account the distance of all the scores from the mean not just the distance between the highest score and the lowest score
255
Evaluation of the mode
+ The mode is not distorted by extreme score called outliers
+ The mode is helpful for discrete numbers; for example it can make more sense to say the average family has two children than 1.89 children
+ Giving the modal group is the only way of giving average of data in categories (e.g. average pet choice)
- There can be no modes if every value is different or multiple modes; this is especially likely in small data sets . This means thin some cases, the mode does not give an exact average value
- The mode does not include all of the values in its calculation so it is not as sensitive as the mean measure of central tendency
256
Evaluation of the median
+ As the median is the central value, its calculation is not affected by extreme outlier scores
+ The median is very easy to calculate
- The median score does not include all of the values in its calculation, so it is not as sensitive as the mean measure of central tendency
- If there are an even number of data points, unlike the mode, the typical value will be a number that is not one of the recorded values
257
Evaluation of the mean
+ All raw data points are used (represented) in calculating the mean. This means the mean is the most sensitive measure central tendency
- Due to the sensitivity of the mean, the mean is distorted by extremely high or low values (outliers)
258
Evaluation of the range
+ The range is easy to calculate, especially compared to the alternative measure of dispersion, standard deviation
- Extreme scores easily distort the value
- The range does not show if the scores are clustered around the mean or more evenly spread out
259
Evaluation of standard deviation
+ SD includes all values in its calculation, making it more sensitive than the range
+ The SD provides information about the spread of scores
- Extreme scores also distort the SD
- The SD is significantly more difficult to calculate than the range
