Paper 2 - Research Methods Flashcards
(111 cards)
1
Q
What is A01?
A
Describe, Outline and Explain
2
Q
What is A02?
A
Link to a scenario
3
Q
What is A03?
A
Evaluation points
4
Q
What is a consent form and what is in it?
A
Given before a study to obtain permission to take part
Aim of study, task/task time, risks, confidentiality, right to withdraw, any questions, place to give consent
5
Q
What is a debrief and what is included?
A
Given added experiment has taken place
Aim/purpose of study, deception (if used) support offered, right to withdraw, confidentiality, any questions
6
Q
What are the 6 ethical issues and what do they mean?
A
Informed consent - Ppts are given comprehensive information in relation to the test to make a decision whether to participate
Deception - Where ppts aren’t told the true aim of the test (only done of necessary and support is available after study)
Right to withdraw - Ppts are free to leave at any point or withdraw data up to 4 weeks after the debrief
Protection from harm - Safety and well-being of ppts is protected at all t8me (psychologically and physically)
Confidentiality - Any information provided must be kept confidential and private
Privacy - Do not invade body (observation) or minds of ppts (questionnaires + interviews)
7
Q
What is the job of the British Psychological Society (BPS)?
A
Oversees psychologists and develops the code of ethics to …
Protect the welfare of ppts / patients
Act as a framework to practise
Promote professional standards
Guide decisions about appropriate conduct
8
Q
What is the ethics committee?
A
Review all research proposals to check it is ethical to conduct or do under the name of psychology
9
Q
What is cost-benefit?
A
The benefits of knowledge gained MUST outweigh the cost (psychological and/or physical damage to ppts)
10
Q
What is control of variables?
A
When doing experiments we must factor in elements to ensure scientific principles are applied and objective, true data is obtained. Variables must be controlled (IV, DV).
11
Q
What is control?
A
Keeping variables constant. Regulated. Factor in potential extraneous variables/something other than the IV that impacts on the DV. (E.g. noise, temperature, keeping conditions the same).
12
Q
What is mundane realism?
A
Does the ‘set up’ of the study (the task) reflect the task in everyday life? If not it backs mundane realism. Really important to try to make the study reflect everyday life.
13
Q
What is generalisability?
A
Can the results of the study apply to 1. Everyday life? (environment) 2. Everyone? (Ppts).
14
Q
What is validity?
A
Measures what it intends to.
15
Q
What is internal validity?
A
Does the test measure what it should?
16
Q
What is external validity?
A
Ecological - can it apply to other settings, Population - can it apply to other people, Temporal - can it apply to different times
17
Q
What is an aim?
A
General statement about the topic you are studying (not testable)
18
Q
What is a hypothesis?
A
A prediction (is testable)
19
Q
What does operationalise mean?
A
Variables must be clear, specific & measurable
20
Q
What does IV mean?
A
The ‘thing’ you change (the 2 conditions)
21
Q
What does DV mean?
A
Measurement
22
Q
What is a 1 tailed / Directional Hypothesis? (experiment + correlation included)
A
One condition will be better than the other
Experiment : 16-18 yr old boys will have bigger hands (cm) than 16-18 yr old girls
Correlation : There will be a positive correlation between height & shoe size
23
Q
What is a 2 tailed / Non Directional Hypothesis? (experiment + correlation included)
A
No past research
Experiment : There will be a significant difference in hand size (cm) between 16-18 yr old males and females
Correlation : There will be a correlation between height and shoe size
24
Q
What are the 4 experimental methods?
A
Laboratory - Conducted in carefully controlled environments
Field - In natural environment, IV manipulated
Natural - Used to investigate casual relationship in situations where IV cannot be manipulated by an experimental environment
Quasi - IV not manipulated & based on pre-existing differences in people(age & gender). DV is naturally occurring (exam results)
25
Strengths & Weaknesses of the 4 experimental methods
L + high controlled environment, infer cause & effect relationship, ethical as ppts know they are taking part, can be replicated
L - artificial situation, lack ecological validity & mundane realism, ppts effect, demand characteristics, low internal validity
F + high ecological validity in real life situations, natural behaviours from ppts
F - less ethical (ppts not alway aware they are part of an experiment
N + research where IV can’t be manipulated for ethical/practical reasons (e.g. studies on privation, so study real problems + increase mundane realism & validity)
N - doesn’t demonstrate casual relationships (IV not directly manipulated), lot of extraneous variables, only used when conditions vary naturally, ppts may know they are studies (ppts effect, investigator effects, demand characteristics)
Q + allow comparisons between types of people, high control
Q - only used when conditions vary naturally, ppts may know studies + not randomly allocated (affects DV) DV may be fairly artificial task, reducing ecological validity
26
What are the 3 experimental designs
How people are used in conditions
1. Independent measures
2. Repeated measures
3. Matched pairs
27
What is independent measures? (+/-)
Different people in each condition (random allocation is essential, e.g. names in hat)
+ No order effects, less chance of demand characteristics, fewer materials
- More ppts needed (than repeated measures), individual differences reduce validity
28
What is repeated measures? (+/-)
Same people in both conditions
+ fewer ppts needed, eliminates individual differences
- Ppts experience order effects & demand characteristics
Way to get around this is counterbalancing (AB/BA) - in experimental design doc
29
What is matched pairs? (+/-)
Ppts matched in a variable pre-test, then each person in the pair is randomly allocated to condition 1 and condition 2 (results compared)
+ No order effects, one set of materials, no demand characteristics
- time consuming, no two people are exactly the same
30
What does extraneous variables mean?
Variables other than the IV, but could affect the results (DV) of the experiment.
31
What are the main 3 extraneous variables?
Participant Variables - How do people differ?
People alter in terms of age, gender, mood and experience. The researcher needs to consider this for ‘control’ to ensure the experiment reduce extraneous variables (changes relate to the research)
Situational Variable - How can a situation differ?
Practice/fatigue effect can occur in repeated measures (control through counterbalancing)
Participant Effects - How can a person affect the study?
Investigator Effects (Cues from the investigator that encourage certain behaviour to the ppts on how to act - a way around this is video record instructions or have written instructions)
Demand Characteristics (Ppts look for cues on how to behave in the study & will respond to these (help or hinder the study) therefore reducing validity.)
32
What are the 3 ways to deal with extraneous variables?
Double Blind - Ppts & researcher doesn’t know what condition they are in (reduces investigator effects)
Single Blind - Ppts ONLY doesn’t know what condition they are in (rescues demand characteristics)
Standardised Procedures - Every ppt does exactly the same thing in the same order. Instructions are pre-recorded or written down.
33
What is confounding variable?
A third Variable : When an extraneous variable have not been properly controlled, it affects the DV (it confounds the results)
34
What is structured observations in relation to observational techniques?
You need to define behaviour categories
These need to be objective, over all possible behaviours and be mutually exclusive (separate)
They must be operationalised (clear, specific and measurable)
They lead to a testable hypothesis
They allow for high reliability
Quantify data = easier to analyse
35
What are the next stages of structured observations in relation to observational techniques? How do you define record/sample?
You then need to define how you will record/sample:
Event sampling : every time the behaviour happens you record
Good for infrequent behaviour (get all behaviour) - complex behaviours are simplified
Time sampling : Reckre behaviours in a given time frame, (e.g. first 10 seconds of each minute
Good for really busy observations, data recorded as intervals
Record using a tally system (increase reliability as can replicate)
36
What are the 6 types of observation?
Overt, covert, participant, non-participant, controlled, naturalistic
37
What does overt observation mean? (+/-)
Ppts are aware they are being studied
+ more ethical (than covert)
- demand characteristics (reduces external validity)
38
What does covert observation mean? (+/-)
Ppts are NOT aware they are being studied
+ demand characteristics are reduced (increases internal validity)
- ethically questionable, privacy may be affected
39
What does participant observation mean? (+/-)
Researcher takes part in their study
+ greater understanding (increases external validity)
- lose objectively, reduces internal validity (Zimbardo)
40
What does non-participant observation mean? (+/-)
Researcher remains separate from the research group being study
+ more objective (than ppt observation)
- loss of insight at researcher is too far removed from the situation (reduces external validity)
41
What does controlled observation mean? (+/-)
Some control/manipulation of variables
+ can be replicated due to standardised procedures
- behaviour may be artificial due to setting (low external validity)
42
What does naturalistic observation mean? (+/-)
Takes place where the target behaviour naturally takes place
+ in a natural context, high external validity (generalisable)
- low control making it hard to identify patterns
43
Self report is …
Subjective, BUT it is the only way to get info about what someone is thinking
44
What are the 2 methods for self report?
Questionnaires
Interviews
45
Explain questionnaires in terms of self report methods and give the 2 types of questions (+/-)
A written method of questions, researcher does not have to be present
Questions need to be clear, concise, unambiguous (not double barrelled)
2 types of questions
Open - ‘Do you like chocolate?’ (+ Lots of detail = gives understanding) (- Harder to analyse than closed)
Closed - ‘Do you like chocolate?’ Yes or No (+ Easy to analyse) (- Lacks detail and understanding)
46
Strengths and weaknesses of questionnaires in relation to self report methods?
Strengths
Collects data quickly, can be replicated, no ethics more data obtained than interview
Weaknesses
Subjective social desirable answers given, people have a tendency to agree even if they don’t
Response set happens, choose all yes/no without reading questions.
47
Explain interviews in terms of self report methods (3 types, +/-, way to solve)
STRUCTURED
A prepared set of questions that will only be used in a fixed order
+ Replicable / Standardised (scientific) / Easy to analyse
- Interviewer expectations may influence
Way to solve = trained interviewer or inter-rater reliability
UNSTRUCTURED
An aim but no fixed questions, the interview determines questions
+ More detail / Can explore interesting things that arise in interview
- Harder to analyse / Requires a trained interviewer to keep focus
SEMI-STRUCTURED
A set of questions to initiate the interview but room for further questions to explore topics that arise in the interview
48
What are 5 other study’s?
Case study / meta-analysis / content-analysis / cross cultural / longitudinal
49
What are case study’s (+/-)?
An in-depth study of a single individual or a small group (5) data gathered from many research methods
+ highly detailed / one is enough to change a theory / high ecological validity
- each case study is unique (results cannot be generalised) / may be researcher bias / time consuming / cannot be replicated
50
What is meta-analysis (+/-)?
This technique produces an effect size of the DV in order to assess overall trends of many studies put together
+ increases validity of conclusion drawn / allows us to reach an overall conclusion by having a statistic to represent the findings
- studies not truly comparable as research designs differ / slightly different hypotheses and different levels of control
51
What is content-analysis (+/-)?
Read the document/interview transcript. Find common themes (give examples) / Identify the behaviour categories / re-read and tally up the behaviour categories (themes)
+ no ethical issues / high ecological validity / can be replicated
- subjective & bias (choose behaviour categories and info can be mis interpreted) / behaviour may be taken out of context
52
What is cross cultural (+/-)?
Comparing behaviours in different cultures. A way to see if certain behaviours are due to nature or nurture
+ explores the nature and nurture debate when explaining human behaviour
- can cause issues such as bias and impose etic
53
What is longitudinal (+/-)?
Study over many years. Comparisons are made at certain milestones/ages in a persons life and conclusions can be drawn.
+ allows to see trend and patterns in behaviour over time
- high drop out rates can skew data
54
What are correlations?
Shows a relationship between 2 variables (good starting point for research)
Does not show causation (cause & effect, cannot explain it)
Represented on a scatter graph (use crosses not dots. DO NOT JOIN THE DOTS)
Look for a co-efficient (e.g. 0.75 strong co-efficient, 0.25 weak co-efficient)
Extraneous variables are not controlled
Also scattergram shows :
Positive correlation - as 1 variables increases, so does the other
Negative correlation - as 1 variable increases, the other decreases
Zero - (no relationship)
55
What words are used in a question when relating to a correlation?
Association
Correlation
Relationship = Scattergram/graph
These words mean it’s a correlations and not an experiment
56
How do you write a hypothesis for a correlation?
1 tailed - There will be a positive relationship between height and shoe size for example
2 tailed - There will be a relationship between liking cheese and height for example
57
What does linear and curvilinear mean in relation to correlation?
Linear - Variables almost always are plotted in a line (positive and negative)
Curvilinear - Relationship is curved and usually predictable (e.g, inverted U (anxiety in memory) another example = time of day and light)
58
Strengths of correlation
Allows to see if two variables are related
Can easily be replicated
It is a good starting point to see if two variables are related then further research can be conducted
Ethical
59
Weaknesses of correlation
Only shows a relationship
Doesn’t show cause and effect (causation - no explanation)
No control of extraneous variables
60
What is reliability?
Replicabilty. Is the tool (measure) consistent. If we repeated the study we should get the same results.
61
What are the 2 ways to test reliability (explain them)?
Test-Retest = Test the person twice! Do test, wait a given time period and then test again. Results from both tests are correlated (looking for a 0.8 co-efficient), this is the easiest method
Inter-rater reliability (questionnaires and interviews) / Inter-observer reliability (observations) = 2 or more observers agree behavioural categories / sit and record separately / at the end they correlate their data / looking for a 0.8 co-efficient
62
What is reliability measured in?
A correlation, to be reliable, data needs to have a co-efficient of 0.8
63
What are the 4 ways to improve reliability (explain them)?
Questionnaires - Rewrite the questions (replace some open questions with closes, fixed choice alternatives which are less ambiguous)
Interviews - Improve training (always try use the same interviewer, use a trained interviewer so that they avoid leading/ambiguous questions (costly)
Observations - Operationalise behaviour categories (behaviour categories must be operationalised (clear, specific and measurable), they must not overlap)
Experiments - Standardise procedures (the procedure must be the same for everyone, every participant does the same thing in the same order)
64
What is validity?
Measuring what you intend to measure
65
What are the 2 types of validity (explain them)?
Internal - Is the tool measuring what it should (are demand characteristics controlled)
External - Is about the results and if they apply to everyday life
ecological = do the results apply to all settings in everday life
population = do the results apply to different groups of people
temporal = can the reiskrs apply over time
66
What are the 2 ways to assess validity (explain them)?
Face Validity - Ask an expert if ‘on the face of it’ your tool measures what you want it to
Concurrent Validity - Using your tool conduct a pilot study & then compare your results to an already established test (same/similar test) & correlate findings. It is valid if you get similar results.
67
What are the 4 ways to improve validity?
Questionnaires - Add lie scales in, ensure confidentiality so people are honest, if only measuring 1 question (add more questions in to reduce demand characteristics)
Interviews - Use the same interviewer or a trained interviewer (they are trained to avoid using ambiguous questions, leading questions and will not give cues
Observations - Behaviour categories must be operationalised (clear, specific, measurable)
Experiments - Use standardised procedures and a control group to compare to
68
What are the types of data
Qualitative
Quantitative
Primary
Secondary
Meta-analysis
69
What is qualitative research/data (from where, how to draw conclusions, +/-)
Qualitative Research - Quality (words), highly detailed and meaningful data
From - unstructured interviews, open questions, case studies and observations
How to draw conclusions - Data needs to be categorised (pre-existing (set before analysis) & emergent (find during analysis). Categories summarise the data and evidenced with quotes from actual research.
Strengths - highly detailed, high validity, meaningful
Weaknesses - hard to replicate, subjective during analysis, time consuming to analyse
70
What is quantitative research/data (from where, +/-)
Quantitative Research - Quantity (numerical data), less meaningful information, has a narrow focus. Qualitative can become quantitative (using content analysis)
From - experiments, correlations, structured observations, structured interviews, closed questions
Strengths - easy to analyse, easy to replicate
Weaknesses - less meaningful, lacks ecological validity
71
What is primary data (+/-)
Researcher collects research first hand
+ Researcher has control over the data, fits the study (get applicable data)
- Very lengthy therefore very expensive (designing and recruiting ppts)
72
What is secondary data (+/-)
Research that already has been collected by another researcher (for their purpose)
+ Easier & cheaper than other data, less time consuming (than primary), such data will have been statistically tested therefore significance is known
- From some studies the data may not fit the needs of the study, data may be outdated
73
What is meta-analysis (+/-)
Lots of studies with similar hypotheses are drawn together and 1 conclusion drawn
+ Increases validity of conclusions, sample is larger than original, increases generalisability
- Publication bias (researcher may not select all relevant studies which will lower validity), many confounding variables (sample size, who the sample are), researcher selects studies (may only choose specific ones)
74
2 types of descriptive statistics
Measures of central tendency
Measures of dispersion
75
Measures of central tendency (what are the 3, method, +/-)
Mean : The average
Method : Add all figures up in a data set and divide by the number of data points.
Strengths : Takes into account all scores
Weaknesses : Skewed by extreme scores (anomalies affect the mean)
Median : Middle value
Method : Put all data in order and identify the middle value
Strengths : Not disturbed by extreme scores, used for ORDINAL DATA
Weaknesses : Not very precise (only tells the middle figure nothing else about data)
Mode : Most occurring
Method : Put all data in order & identify the one (or more) that occurs most
Strengths : Used for NOMINAL DATA, tells frequency
Weaknesses : Not always a mode, sometimes may modes
76
Measures of dispersion (what are the 3, method, +/-)
Range : Shows range of scores
Method : Highest - smallest figures
Strengths : Easy to calculate
Weakness : Could be affected if highest or lowest figure is extreme
Standard Deviation : Spread of scores around the mean
Strengths : The most precise measure of dispersion
Weaknesses : Only used for INTERVAL DATA, harder to calculate
77
Representing Data : what are examples of tabular displays and graphical displays
Tabular = summary tables
Graphical (Numerical data can be visually displayed) = bar charts / scatter grams / histograms
78
Representing data : explain summary tables
Tables must always be:
◦ Fully labelled (x and y axis)
◦ Have a clear title
◦ Column headings
The title must have reference to the units of measurement, the numbers in the table, as well as the two conditions of the experiment.
A verbal summary follows the table
79
Representing data : explain bar charts
Used to show frequencies in columns with the frequency on the vertical axis and the variable on the horizontal axis. When columns are drawn there is always a gap.
80
Representing data : explain scattergrams
Used for correlational data. One variable is plotted on each axis with each pair shown as a single cross/dot. The pattern of dots indicates a positive, negative or zero correlation. NO LINE
81
Representing data : explain histograms
Similar to a bar chart except that the area within the bars must be proportional to the frequencies represented. Represents continuous data - age, temp, time
Used for interval data. No gaps between the bars.
82
Representing data : normal distribution
Used to see how traits are distributed through a population. “bell curves” - tracks rare occurrences of a trait on both the high and low ends of the “curve” with the majority of occurrences appearing in the middle section of the curve.
E.g. IQ tests with the majority of the population scoring within the “normal” or middle-range of intelligence.
83
Representing data : skewed distribution
Sometimes scores are not distributed equally around the mean.
i.e. in a test of depression 0-50 represents normal behaviour, 50+ represents clinical depression. If we plot the distribution scores for 100 people we would expect most scores to be towards the low end rather than the high end of this score range – produces a positive skewed distribution. Negative skewed distribution might happen for a very easy exam so a lot of people get the high mark, not many getting the low mark.
84
What is the sign test used to work out?
The only inferential stat test you work out!
Used to work out if your study is significant (based on data collected)
85
Sign Test : Levels of Measurements - What ‘level’ is your data
Nominal - Categorised data (smoker/non-smoker, music genre, dream type)
Ordinal - Subjective scores, data in order (how well did you sleep 1-10)
Interval - Objective universal measurements, where there are equal increment/intervals between scores (mm, cm, m, km, temp)
Ratio - True zero point as is most measures of physical quantities
86
When to use the sign test
1. The experimental design is repeated measures (ppts do both conditions)
2. On ‘test of difference’ studies
87
Sign Test : How to work out the calculated value of ‘s’
Take away column A from column B (or other way round makes no difference)
State the sign (either ‘+’ or ‘-‘ for each ppt)
Any ‘0’ eliminate from the sample data (your ‘n’ number will decrease, n = number of ppts)
Add up how many ‘+’ signs you have
Add up how many ‘-‘ signs you have
S = the number of the least frequent sign
88
Sign Test : How to work out if your study is significant
*Get the calculate value of S - you have to work out!
Using the critical value table:
1. Work out if it’s a 1 tailed (directional) or 2 tailed (non-directional). Normally clear from scenario or based on the hypothesis you have written or it will ‘state in the question - it will be clear’.
2. Use the column that states ‘0.05’ level of significance - unless question states otherwise.
3. Get ‘n’ number (number of ppts), remember to eliminate any ‘0.5’ use this row.
4. Based on point 3 and 4 you should have your ‘critical value’.
5. Under the critical values table it will tell you if your calculated value of 5 needs to be greater/equal to/less than the critical value (from table) to be significant.
89
Inferential Statistics : Significance in Psychology
Always use a 0.05 level of significance in psychology unless the Q states otherwise.
0.05 means you are 95% sure the IV manipulation caused the change in DV.
OR
5% that the effect occurred by chance (extraneous variables, chance/something not controlled well)
90
Inferential statistics : What are the levels of measurement and define them
Nominal: Numerical data from categories (I.e. smoker = 400, non smoker = 793)
Ordinal: Data ordered in some way, subjective measurement (I.e. how well you slept in a scale of 1-10)
Interval: Equal increments between each unit, objective and universal (I.e. time, temp, distance, volume)
Ratio: there is a true zero point as in most measures of physical quantities.
91
Inferential statistics : how to choose a test
Is it a test of difference or relationship/correlation/association (one of these 3 words will be in the scenario if it’s a relationship)
What is the level of measurement (data)
What are the experimental design (how are ppts used: repeated or independent measures?)
92
Inferential statistics : what are the 8 tests?
Lesson in inferential statistics explain what test of, level of measurement and experimental design
Sign test
Chi square
Spearman’s R
Pearsons R
Wilcoxan
Man Whitney U
Related T Test
Unrelated T Test
93
Critical value, how to check if a study is significant
1. You need the observed/calculated value (in the question)
2. The level of significant (usually 0.05 unless told otherwise)
3. 1 tailed or 2 tailed (doesn’t always state in the question, go back to the scenario. If it doesn’t state, it doesn’t matter which one you use)
4. The critical value (you find in the table)
5. To look under the table to know which way round it has got to be
94
Critical value, how to write a conclusion
1. As the Calculated/observed value (GIVEN TO YOU IN THE QUESTION)
2. Is more than/less than the Critical value (IN THE TABLE)
3. Test is significant/not significant
4. For a “One/two tailed test”
5. At a ? Significance level (p=0.05)
6. For “No. of participants (n=?)”
7. Therefore the hypothesis/Null hypothesis should be accepted
95
Critical value
Inferential statistics lesson in docs
96
What is a type 1 error
Type 1 = null rejected when it’s true, too optimistic. Significant level too lenient (0.10) false positive.
97
What is a type 2 error
Type 2 = null accepted when false, pessimistic. Significance level too strict (0.01).
98
Reporting psychological investigations : abstract
Abstract = 200 word summary of the study, reporting aim, methods, results, conclusion.
99
Reporting psychological investigations : introduction
Introduction = Review of literate in the specific topic that leads to the formulation of your hypothesis.
100
Reporting psychological investigations : method
Method = Design, sampling, materials, procedure, ethics, how to ensure relatability and validity.
101
Reporting psychological investigations : results
Results = Descriptive and inferential statistics (stat tests)
102
Reporting psychological investigations : discussion
Discussion = Summary, links to earlier research, limitations of research, implications and possible future research.
103
Reporting psychological investigations : referencing
Referencing = book - Jones, R. (2001) The title. London : books
Journal - Jones, R. (2001) the title. Journal name, 4 (1), 23-27
104
What is a peer review and +/-
Needed for allocation of research funding.
Validation of the quality and relevance of research.
Allows improvements and amendments to be made.
+ Improves quality of research being published
+ Increases probability of weaknesses/erros being identified
+ Reputation of psychology as a science high
- Finding an expert
- Publication bias if ‘juicy’ research
- Cannot change already published research
105
Define replicability
Replicability = Findings repeatable, generalisable
106
Define theory construction
Theory Construction = General law, develop and test hypothesis
107
Define deduction
Deduction = Developing new hypotheses from existing theories
108
Define paradigm shift
Paradigm Shift = A scientific revolution, a new way of thinking due to research
109
Define empirical methods
Empirical Methods = data collected through direct experience (rather than just logic)
110
Define falsifiability
Falsifiability = Must be able to be proved false (can re test)
111
Define objectivity
Objectivity = fact, unbiased
