AQA Psychology A Level - Research Methods

Define IV (1)

Define DV (1)

• what must these both be? (1)

• what does it mean to standardise procedures? (1)

IV - variable manipulated by researcher to see effect (has two or more conditions/levels)

DV - variable that is measures to gain results

→ both should be operationalised (so easy to test/measure)

Standardised procedures:

• ppts do same thing for each condition/identical method for ppts in each condition

Define hypothesis

Define null hypothesis

Define alternative hypothesis

Hypothesis - prediction that includes all conditions of IV and DV (states relationship between IV and DV)

Null hypothesis - prediction of no effect/relationship between IV and DV

Alternative hypothesis - prediction of effect/relationship between IV and DV

Define directional hypothesis (3)

Define non-directional hypothesis (3)

Directional hypothesis:

• states what direction results will go in (e.g. increase or decrease)

• based on previous research with similar results

• one-tailed

Non-directional hypothesis:

• doesn’t state directional results will go in, but does state effect (e.g. there will be a difference between results)

• is not based on previous research or conflicting results

• two-tailed

Difference between confounding and extraneous variable?

Extraneous - is not related to what is being studied, but can have an affect on the outcome/DV

Confounding - type of extraneous variable that is related to the IV, and has an effect on the outcome/DV

Define ‘mundane realism’ (1)

Define ‘generalisation’ (1)

Mundane realism - the extent to which an experiment mirrors the real world

Generalisation - applying the results of research to everyday real life (if there is mundane realism)

Define ‘validity’ (1)

• describe the two types of validity (3 points for each)

Validity = how well a study/research measures what it says it measures

Internal validity:

• considers effect of confounding/extraneous variables on effect on DV

• considers if aim of experiment was tested

• considers if study had mundane realism

External validity:

• ecological validity - extent to which research findings can be generalised to other/real life settings

• population validity - extent to which research findings can be generalised to other people/samples

• temporal validity - extent to which research findings can be generalised to other time periods

Define ‘confederate’ (1)

Define ‘pilot study’ (1)

• why are pilot studies used? (1)

Confederate - researcher/a person who is aware of the aims of the study and is playing another role in the study (not naive ppt)

Pilot study - small-scale trial run of study before real thing

• can determine if anything needs to be adjusted before real study (e.g. questions in interview) - so time and money not wasted on real study

Name and describe the 4 experimental methods (2 points for each - setting and IV manipulation)

1. Lab - controlled setting, IV manipulated

2. Field - natural setting, IV manipulated

3. Natural - natural setting, naturally occurring IV

4. Quasi - generally controlled setting, IV is a variation in sample/existing difference in ppts

Advantages and disadvantages of each experimental method

• lab (3 advantages, 2 disadvantages)

• field (2 advantages, 3 disadvantages)

• natural (2 advantages, 3 disadvantages)

• quasi (1 advantages, 3 disadvantages)

Lab:

+ reliable/can be replicated (standardised procedures)

+ lacks extraneous variables (control over variables/standardised procedures)

+ can establish cause and effect

- lacks validity

- artificial

- causes demand characteristics

Field:

+ some reliability

+ some validity

- some demand characteristics

- lacks some reliability

- lacks some validity

Natural:

+ high validity

+ low demand characteristics

- lack reliability

- extraneous variables

- cannot establish cause and effect

Quasi:

+ allows comparisons between people

- can only be used when conditions vary naturally

- lack ecological validity

- demand characteristics

Define ‘true experiments’ (3)

• therefore, which experiments are considered ‘true’ (2)

True experiments:

• when there is a control group and an experimental group

• ppts are randomly assigned

• researcher can manipulate variables (IV, confounding, etc.)

→ lab and field experiments are considered ‘true experiments’

Name and describe the 3 experimental designs

Independent Groups – each group completes a different condition.

Repeated Measures – all participants completes all conditions.​

Matched Pairs Design – each groups is made up of participants who have been matched on relevant characteristics, e.g. IQ​

Advantages and disadvantages of each experimental design

• independent groups (4 advantages, 2 disadvantages)

• repeated measures (2 advantages, 4 disadvantages)

• matched pairs (5 advantages, 3 disadvantages)

Independent groups:

Advantages	Disadvantages
+ can use the same resources + can compare results quickly + lacks demand characteristics	- participant variables - need more participants

Repeated Measures:

Advantages	Disadvantages
+ no participant variables + need less participants​	- need different resources for the conditions - demand characteristics - order effects - can be time consuming

Matched Pairs:

Advantages	Disadvantages
+ less participant variables + lack demand characteristics + no order effects + can make quick comparisons of results + can use same resources​	- no two people are exact - need more participants - gathering sample can be time consuming

Give way of overcoming limitations for each experimental design and explain how it overcomes limitation

• independent groups (1)

• repeated measures (1)

• matched pairs (2)

Independent groups:

• Can use random allocation or matched pairs design so ppt variable distributed evenly

Repeated measures:

• Can use counterbalancing for order effects

Matched pairs:

• Restrict matching criteria (easier to match), and conduct a pilot to consider key characteristics before study

What is counterbalancing used for (1)

• what are the two methods of doing this?

  • method 1 (2)

  • method 2 (5)

Counterbalancing - reduces order effects as all ppts experience conditions of IV in all orders

Method 1: AB or BA

• each ppt in group 1 does condition A, then B

• each ppt in group 2 does condition B, then A

Method 2: ABBA

• all ppts do condition A and B twice in opposite/different orders:

  • first, they all do condition A (trial 1)

  • then, they all do condition B (trial 2)

  • then, they do condition B again (trial 3)

  • finally, they do condition A again (trial 4)

Difference between a ‘population’ and a ‘sample’?

Population - the whole group of people with the characteristic to be studied

Sample - subset/group of the population which are selected for the study

Name the 5 types of sampling

1. Opportunity

2. Random

3. Stratified

4. Systematic

5. Volunteer

Describe how each sampling method is carried out

• opportunity (2)

• random (2)

• stratified (2)

• systematic (2)

• volunteer (2)

Opportunity sampling:

1. select people out of population who are most easily available and willing to take part

2. e.g. people walking by in street, people at school

Random sampling:

1. use random allocation technique to gather enough ppts for sample e.g. putting all names in a hat and pulling them out at random, random number generator

2. each ppt has an equal chance of being selected

Stratified sampling:

1. divide population into subgroups e.g. boys and girls, age groups, etc.

2. randomly select ppts from each subgroup (e.g. names in hat) until enough are selected for sample/each condition

Systematic sampling:

1. use predetermined system to select every nth person from population until enough for sample

2. e.g. randomly select first person, then select every other person onwards

Volunteer sampling:

1. advertise chance to take part in study e.g. online, noticeboard

2. sample select themselves/willingly choose to take part

Advantages and disadvantages of each sampling method

• opportunity (1 advantage, 2 disadvantages)

• random (1 advantage, 2 disadvantages)

• stratified (1 advantages, 2 disadvantages)

• systematic (2 advantages, 1 disadvantage)

• volunteer (2 advantages, 1 disadvantage)

Opportunity:

+ quick and easy

- unrepresentative of target population (e.g. only selecting people on a Monday may exclude those not working)

- potential investigator bias

Random:

+ lack of investigator bias (all ppts have equal chance of being selected)

- unrepresentative of target population

- may be time consuming (gather list of whole population)

Stratified:

+ representative of target population

- time consuming

- potential investigator bias

Systematic:

+ is list is random no investigator bias

+ quick and easy

- unrepresentative of target population

Volunteer:

+ willing and motivated ppts

+ access to variety of ppts, so maybe more representative

- volunteer biased sample e.g. only motivated ppts, only those with free time

Difference between time and event sampling

Time sampling - recording (e.g. tallying) how often a certain behaviours are seen in specific time intervals

Event sampling - recording every instance of a specific behaviour

Name and describe the 4 BPS guidelines

1. Respect - keep confidentiality, get informed consent, do debriefing for ppts dignity/worth

2. Competence - research qualified and keeps high standards

3. Responsibility - duty to psychology as a science, ppts right to withdrawal, protecting ppts e.g. debriefs

4. Integrity - no deception in reporting findings to respect psychology as a science

Name the 6 ethical issues in the BPS guidelines and briefly describe them

1. Informed consent → ppts given comprehensive info about purpose and nature of study and their role so they can make informed decision to take part

2. Right to withdraw → ppts can stop participating in study if uncomfortable and can withdraw their data

3. Deception → ppts not told true aims of study, so can’t give true informed consent

4. Protection from harm → ppts shouldn’t receive negative physical/psychological effects e.g. injury, embarrassment

5. Confidentiality → ppts personal info is protected and secure

6. Privacy → ppts right to control flow of info about themselves

Give ways of overcoming each ethical issue

• informed consent (2)

• right to withdraw (1)

• deception (1)

• protection from harm (3)

• confidentiality (1)

• privacy (2)

1. Informed consent:

• have ppts sign document agreeing to participate in study, knowing the nature/purpose of it and their role and their right to withdraw

• if ppts are children under 16, need parents/guardian signature

2. Right to withdraw:

• ppts informed at start and end of study about their right to withdraw from the study and their data

3. Deception:

• ppts fully debriefed after study so they can ask questions and withdraw any data

4. Protection from harm:

• avoid any risks greater than those experienced in everyday life

• stop study if harm suspected

• give right to withdraw and offer counselling where needed

5. Confidentiality:

• maintain anonymity e.g. use initials/numbers instead of names

6. Privacy:

• gain informed consent (unless in public space and public behaviour)

• do pilot study to determine any privacy risks

Give the limitations for each ethical issue/how it’s dealt with

• informed consent (2)

• right to withdraw (2)

• deception (1)

• protection from harm (1)

• confidentiality (1)

• privacy (1)

Informed consent →

• giving full info may invalidate purpose of study/increase demand characteristics

• presumptive consent - what people expect may be different from actually experiencing it

Right to withdraw →

• ppts may not withdraw for fear of ‘spoiling study’

• may feel unable to withdraw if being paid

Deception →

• debriefing cannot undo psychological harm

Protection from harm →

• harm may only be apparent after the study/with hindsight

Confidentiality →

• sometimes complete confidentiality not possible e.g. geographical location of school may give away ppts

Privacy →

• no universal agreement about what a ‘public place’ is

Define socially sensitive research

• any research that might have direct social consequences for ppts or the group they represent

Name the other considerations Sieber and Stanley proposed should be taken into account to avoid socially sensitive research (6)

• briefly describe what each means

1. Valid methodology → scientists may be aware of poor methodology in research (which affects findings), but the public and media don’t know.

2. Equitable treatment → any helpful resources should be available/provided to all groups (e.g. treatment for certain disorders shouldn’t be kept from control placebo group).

3. Scientific freedom → scientist is free to study an area but without harm to ppts or institutions in society.

4. Ownership of data → some data may only be owned by research sponsors (e.g. university) while others are available to public to comment on it.

5. Values → issues with clash of values between scientists and funders/recipient of research.

6. Risk/benefit ratio → risks or costs should be minimised, but problems in determining an justifying risks and benefits.

Define ‘demand characteristics’ (3)

Demand characteristics:

• ppts guessing the aim of the study

• therefore acting in a way unnatural to their normal behaviour - may either want to go against or follow ‘expected’ behaviour

• a type of extraneous/confounding variable

Define ‘investigator effects’

Investigator effects:

• cues (other than IV) from investigator that encourages certain behaviour in ppts (intentional or unintentional) e.g. leading questions

• leads to fulfilment of investigators expectations

• a type of extraneous/confounding variable

What are indirect investigator effects? (1)

• give an example

What are investigator loose procedure effects? (1)

• give an example

Indirect investigator effects - (intentional or unintentional) when experiment is designed in certain way to encourage certain ppt behaviour

• e.g. study has limited duration so desired result more likely

Investigator loose procedure effects - (intentional or unintentional) when investigator doesn’t use standardised procedures correctly so there is room for result to be influenced by ppt or another experimenter

• e.g. a separate interviewer in a study doesn’t ask questions in the correct manner

How can demand characteristics and investigator effects be overcome? (3)

• briefly describe each method (1 point each)

• how is each useful to prevent these effects (1 point each)

1. Single blind trial - ppts not aware of research aims and/or which condition of experiment they are in e.g. placebo or real treatment

   • so less likely to guess aims of study

2. Double blind trial - ppts and researcher unaware of which conditions of experiment ppts are in

   • so ppts less likely to guess aims of study and investigator can’t give cues about what they expect

3. Experimental realism - when research task is sufficiently engaging enough so ppts pay attention to task and not the fact they are being observed

   • so less likely to guess aims of study and pick up on investigator cues

Name and briefly describe the 8 main types of observation methods (1 point for each)

1. Naturalistic = natural situation, no manipulation of IV

2. Controlled = likely controlled setting, manipulation of IV

3. Participant = researcher is part of group observed

4. Non-participant = researcher observes from distance

5. Overt = ppts unaware they are being observed

6. Covert = ppts aware they are being observed

7. Structured = researcher uses behaviour categories and sampling procedures to observe

8. Unstructured = researcher observes everything

Advantages and disadvantages of each observation method

• naturalistic (2 advantages, 3 disadvantages)

• controlled (3 advantages, 2 disadvantages)

• participant (1 advantage, 2 disadvantages)

• non-participant (2 advantages, 1 disadvantage)

• overt (1 advantage, 1 disadvantage)

• covert (1 advantage, 1 disadvantage)

• structured (2 advantages, 1 disadvantage)

• unstructured (2 advantages, 2 disadvantages)

1. Naturalistic

+ lacks demand characteristics

+ has ecological validity

- extraneous variables not controlled, so difficult to repeat

- difficult to establish cause and effect

- may never see intended behaviour

2. Controlled

+ lacks extraneous variables#

+ cause and effect can be established

+ reliable, so can be repeated

- lacks ecological validity (low mundane realism)

- hawthorne effect = ppts change behaviour to fit environment (unnatural)

3. Participant

+ more detailed insight into behaviour, so valid

- may create observer bias

- difficult to take notes

4. Non-participant

+ easy to take notes

+ more objective observing

- lacks personal insight/understanding into behaviour

5. Overt

+ easy to gain consent

- hawthorne effect = ppts change behaviour to fit environment (unnatural)

6. Covert

+ less demand characteristics, so natural behaviour (valid)

- consent is hard to get (ethical issue)

7. Structured

+ can focus to record all of a particular behaviour (event) or all of what occurs in a certain time period (time) - straightforward method

+ systematic data and procedure which is easy to repeat

- may lose some details/something important

8. Unstructured

+ useful as pilot study in new research, so expected results gained

+ more detailed, qualitative data

- difficult to pick up on everything

- greater risk of observer bias, may only record behaviour that seems eye-catching

Define inter-observer/rater reliability (2)

• what can increase inter-observer reliability (2)

Inter-observer/rater reliability:

• when reports of two observers/judges watching same event match/correlate

• looking for co-efficient that is 0.8 or above

→ having more than one or two observers

→ having good behaviour categories (standardised)

Name the 2 most common self-report techniques

1. Questionnaires

2. Interviews

Difference between open and closed questions?

Open = doesn’t have fixed answer, so produces qualitative behaviour

Closed = offers fixed number of responses, so produces quantitative behaviour

Difference between structured and unstructured?

Structured = has pre-determined, mostly closed questions

Unstructured = no pre-determined questions, more like a conversation based on a topic, open questions

What makes a good questionnaire/interview? (6)

1. questions should be clear/unambiguous

2. non-leading questions

3. use of filler questions so aim of study not guessed

4. start with easy questions

5. use of positively and negatively worded questions

6. use of a pilot study beforehand

Advantages and disadvantages of self-report techniques

• questionnaires (4 advantages, 4 disadvantages)

• interviews

  • structured (4 advantages, 3 disadvantages)

  • unstructured (3 advantages, 4 disadvantages)

Questionnaires:

+ easy to analyse data and use statistical tests

+ less researcher bias, as researcher usually not present

+ can cover a large sample/large amount of data

+ cost effective

- respondents may not understand questions

- social desirability bias

- specific choices maybe not available, so limits data

- only considers specific time period

Interviews:

→ Structured:

+ easy to analyse/compare

+ less researcher bias

+ can cover large sample/large amount of data

+ easily repeated (standardised)

- respondents may not understand questions

- social desirability bias

- specific choices maybe not available, so limits data

→ Unstructured:

+ detailed data

+ interviewer can explain questions/change wording/as follow up/create new questions

+ may be more valid if built up good rapport

- difficult to analyse

- small samples/less data collected

- investigator bias (e.g. way in which questions asked), so need skilled interviewer

- social desirability bias

Difference between primary and secondary data? (2 points for each)

→ Primary data:

• gathered directly/first-hand from the participants

• is specific to the aim of the study

→ Secondary data:

• has previously been collected by a third party (another researcher or an official body), then used by researcher

• not specifically for the aim of the study

Define ‘meta-analysis’ (1)

• advantages (1) and disadvantages (2)

Meta-analysis: comparison of a number of studies focused on one area to reach an overall conclusion

Advantages - quick away of reaching conclusion

Disadvantages - original studied may not be reliable, or may not concern exact areas wishing to be studied

Define ‘content analysis’ (1)

• explain how this is carried out (3)

• advantages (3) and disadvantages (2)

Content analysis: a technique for analysing data according to themes or categories

1. create relevant categories of behaviour/themes in data (coding)

2. count/tally number of times each category/theme is seen (quantitative data)

3. compare categories/themes results to determine or describe any changes/patterns over course of study (qualitative data) - thematic analysis

Advantages:

+ higher ecological validity - based on what people actually do

+ reliable - can obtain same material and do same analysis

+ can produce quantitative and qualitative data

Disadvantages:

- difficulty deciding categories

- observer bias may have different interpretations of data (especially between cultures)

Define ‘thematic analysis’ (1)

• explain how this is carried out (4)

• advantages (1) and disadvantages (1)

Thematic analysis: a method of analysing qualitative data by searching for common themes/categories

1. initially analyse the data by re-reading it multiple times

2. break it down into meaningful/relevant codes/themes in order

3. review the data with these codes to find overall patterns/themes (qualitative data)

4. use of other existing data to support conclusions

Advantages:

+ summarises data into themes

Disadvantages:

- lengthy process/time consuming

What is meant by a ‘case study’? (5 points)

Case Study:

1. an in-depth/detailed study of a single individual, a group of people, or an event over time

2. generally longitudinal

3. usually carried out in the real world using various techniques e.g. observation, testing, interviewing

4. idiographic and very individualistic

5. uses scientific methods that aim to be systematic and objective

Give an example of a case study (in memory topic)

HM case study - hippocampus removed to prevent further epileptic seizures, resulting in problems forming new LTM’s

Advantages (3) and disadvantages (4) of case studies

Advantages:

+ gives rich in-depth, detailed data that otherwise be overlooked

+ can be used to investigate rare behaviours

+ if longitudinal, can see changes over time

Disadvantages:

- difficult to generalise findings as individuals/situations are unique (especially for brain damage)

- info gathered often based on retrospective data, so maybe not accurate/hard to verify

- very difficult to replicate case study, so lack reliability

- researcher may develop personal/close relationship with individual, so loss of objectivity

Name and define each measure of central tendency (3)

Mean → add up all values and divide by number of values

Median → order from lowest to highest and find the middle number

Mode → most common value in data set

Advantages and disadvantages of each measure of central tendency

• mean: advantages (1), disadvantages (2)

• median: advantages (1), disadvantages (1)

• mode: advantages (2), disadvantages (1)

Mean:

Advantage - takes into account all values/scores

Disadvantages - can be skewed by extreme scores/anomalies, can’t be used for categorical data

Median:

Advantage - excludes extreme scores/anomalies

Disadvantage - doesn’t take into account all values/scores

Mode:

Advantages - can be used for categorical data, excludes extreme scores/anomalies

Disadvantage - not useful when there are several modes

Name and define each measure dispersion (2)

Range → highest value subtracting/take away lowest

Standard deviation → the average distance of the data item from the mean

Advantages and disadvantages of each measure of dispersion

• range: advantages (1), disadvantages (1)

• standard deviation: advantages (2), disadvantages (1)

Range:

Advantage - easy to calculate

Disadvantage - doesn’t account for extreme extreme scores or distribution of values

Standard deviation:

Advantages - precise as all values taken into account, easy to calculate with calculator

Disadvantage - may still hide extreme values

Define what is meant by ‘correlation’ (1)

• give the key features of a correlation graph (5)

• advantages (2) and disadvantages (2)

Correlation - the relationship between two continuous variables (co-variables)

1. plotted on scattergram

2. each axis has each co-variable (independent variable on x-axis, dependent variable on y-axis)

3. can be perfect positive (+1.0) or negative (-1.0) correlation, or no correlation (0.0) - result of Spearman’s rho or Pearson’s r value

4. has straight line of best fit (if correlation shown)

5. doe NOT show causal relationship/cause and effect

Advantages:

+ shows relationship between two co-variables, which can be further studied

+ secondary data can be used, so less time consuming

Disadvantage:

- doesn’t show cause and effect between co-variables

- other untested variables may be cause of relationship

Define quantitative data

• advantages (3) and disadvantages (2)

• numerical data e.g. tally, range, count

Advantages:

+ easier to analyse and draw conclusions from (e.g. in measures of central tendency, statistical tests)

+ objective measurement, standardised

+ less open to bias

Disadvantage:

- doesn’t give detailed info

- reductionist

Define qualitative data

• advantages (2) and disadvantages (3)

• non-numerical data e.g. words, images

Advantages:

+ gives much more in depth, rich, detailed info

+ not reductionist

Disadvantage:

- harder to analyse

- more subjective, not standardised

- more open to bias

What is meant by normal distribution? (4)

Normal distribution:

1. bell-shaped curve on graph

2. mean, median, and mode at same point

3. symmetrical scores around the mid-point

4. indicates a representative sample e.g. in IQ scores

What is meant by positive skew? (2)

Positive skew:

• median and mode lower than mean

• tails off to the right

What is meant by negative skew? (2)

• median and mode are higher than mean

• tails of to the left

Define ‘reliability’ (1)

Reliability - how consistent a test or study is in producing the same results when repeated under identical conditions

How to improve reliability? (6)

1. inter-observer/interviewer/rater reliability used (multiple observers/interviewers/judges)

2. behaviour categories are operationalised and practiced before study

3. reduce ambiguity in questions e.g. wording

4. test-retest reliability used (repeat study with same ppts)

5. pilot study

6. standardised procedures e.g. method of measuring DV

Define ‘validity’ (1)

Validity = how well a study/research measures what it says it measures, how true/legitimate the observed effect of a study is

Name all the types of validity (7)

• Internal validity

• External validity

• Population validity

• Temporal validity

• Ecological validity

• Face validity

• Concurrent validity

Name the concerns associated with internal validity (5)

1. Investigator effects

2. Demand characteristics

3. Confounding variables

4. Social desirability bias

5. Poorly operationalised behaviour categories

Define ‘face validity’ (1) - give an example

Face validity - the extent to which a test/measure appears to assess what it aims to measure → e.g. a questionnaire investigating anxiety having questions related to feeling anxious = high face validity

i.e. if a test looks like its measuring the intended concept, it has high face validity

Define ‘concurrent validity’ (1) - give example

• when is there high concurrent validity (1)

Concurrent validity - the extent to which there is close agreement between the results of a test being investigated compared to an already established test (e.g. school assessment compared to a standard IQ test)

• if correlation between the test results is over +0.8, there is high concurrent validity

(comparing current method with a validated method to compare results)

Name the features of science (7)

1. Empirical method

2. Objectivity

3. Replicability

4. Theory construction

5. Hypothesis testing

6. Falsifiability

7. Paradigms

Define ‘empirical method’ (1)

• give an example (1)

• info gathered through direct observation or experiment (used to separate unfound beliefs from real truths)

• e.g. and advertisement vs a real product = empirical

Define ’objectivity’ (2)

• not affected by researcher expectations, so free from bias/distortion

• involves systematic, controlled data

Define ‘replicability’ (1)

• when is there high replicability (1)

• difference between reliability and replicability (1)

• repeating experiment using different group of people to confirm outcomes

• procedure is able to be repeated exactly to get same results = high replicability

• reliability - same people tested in same way (e.g. personality quiz repeated by same person), replicability - different people tested in same way (e.g. experiment repeated with different ppts)

Define ‘theory construction’ (2)

• an explanation of facts and observations, which is refined to be suited to new evidence over time

• scientists use inductive and deductive methods so theory comes before or after hypothesis testing

Define ‘hypothesis testing’ (1)

• a good theory must generate testable expectations so it can be checked and modified

Define ‘falsifiability’ (1)

• what is the ‘Falsification Principle’, who created it

• give an example

Falsifiability = the possibility that a statement or hypothesis can be proven wrong

• Popper created Falsification Principle - we cannot confirm a theory, so instead we try to disprove it with even a single observation

• Example: no number of white swan sightings indicates that black swans don’t exist - the theory all swans are white is falsified with one black swan sighting

Define ‘paradigm’ (1)

• who supported this (1)

• what is a paradigm shift (1)

Paradigm = a unified set of assumptions and methods that are currently held about a theory or research

• Kuhn states scientific knowledge develops through revolutions

• Paradigm shift - when theories are continuously refined until a different view of a concept becomes commonly held

What is peer review? (1)

• what is its purpose (4)

Peer review = assessment by other independent experts in the field

Purpose of peer review:

• to validate quality and relevance of research reports (e.g. suitability for publication, methodology, statistics, and conclusions)

• to suggest improvements and prevent errors to research reports/academic journals

• to allocate research funding appropriately

• to assess ratings of university departments

Evaluation of peer review

• strengths (3)

• weaknesses (4)

Strengths:

+ anonymous to gain honest review from fellow experts

+ findings should be novel, interesting and relevant, so add to the knowledge of a particular area

+ helps ensure poor quality work isn’t published in reputable journals

Weaknesses:

- publication bias → towards prestigious researchers/research departments or competitors
- publication bias → towards positive, eye-catching findings/headlines
- publication bias → towards 'established' research areas/status quo (so novel/unusual research hard to publish)
- time consuming and expensive - peer review can take months/years so delays publication of important finding and experts not easy to find

Give examples of how psychology can/has impacted the economy (3)

• Attachment - role of the father → paternity leave

• Psychopathology - treatment of illnesses/disorders → less people ill, so can work

• Stress - finding ways to manage stress → more people working

Name the steps in an Investigation Report in order

1. Abstract

2. Introduction

3. Method

4. Results

5. Discussion

6. References

   → then it is peer reviewed

Describe each section of an Investigation Report

• abstract (2)

• introduction (2)

• method (2) - briefly describe each subsection (5)

ABSTRACT

• 150-200 word summary of whole article/study

• includes aims, hypothesis, method procedures, results, conclusion, and implications of study

INTRODUCTION

• review of previous research to show reasons for research

• includes aims and operationalised null and directional or non-directional hypothesis (depending on trend in previous research)

METHOD

• detailed description of what researcher did (past tense), enough so precise replication would be possible

• includes design (e.g. repeated measures, covert, IV and DV + justification), participants (sampling method justification and sample info e.g. age, how many), apparatus/materials, procedures (standardised instructions, setting, order of events), and ethics (issues and how overcome) → all in prose, no subheadings

Describe each section of an Investigation Report

• results (2)

• discussion (2) - subsections (4)

• references (2) - briefly describe the general format (2)

RESULTS

• descriptive statistics (tables, graphs, measures of central tendency, dispersion) and inferential statistics (statistical tests, calculated values, significance levels)

• includes qualitative research themes and categories

DISCUSSION

• interpretation/conclusion of results (which hypothesis accepted/rejected)

• includes implications for future research, relationship to previous research, consideration of methodology (criticisms, improvements), and suggestions for future research

REFERENCES

• all journal articles and books used

• general format:

  • Journal: Author name(s), date, title of article, journal article, volume, page numbers

  • Book: Author name(s), date, title of book, place of publication, publisher

Name and define the three levels of measurement of a study/data

• give an example for each

1. Nomimal data - separated in categories e.g. grouping ppts into small, medium, tall

2. Ordinal data - ranked in order without equal intervals e.g. grouping ppts in order of height

3. Interval data - in equal intervals e.g. ppts rank happiness on scale of 1-10

Criteria for a parametric vs non-parametric test (3 each)

Parametric:

• DV is a safe measure (can be objectively, reliably measured e.g. bpm, test score)

• Interval data

• Normal distribution (characteristics that cluster around the mean e.g. height, IQ, shoe size)

Non-parametric:

• DV is an unsafe measure (is subjectively, less reliably measured e.g. rating stress on questionnaire)

• Nominal data

• Ordinal data

Name the 8 statistical tests

1. Mann-Whitney

2. Wilcoxon

3. Spearman’s Rho

4. Pearson’s R

5. Related-T

6. Unrelated-T

7. Chi Square

8. Sign test

When choosing a statistical test, what 3 aspects should you consider?

1. Parametric or non-parametric (and so which data - internal, nominal, or ordinal) - is DV safe or not

2. Looking for a difference/association or a relationship/correlation

3. Uses repeated measures or independent groups design (repeated is related data, independent is unrelated data)

Name the characteristics of each statistical test (3 characteristics for each, 8 tests in total)

Mann-Whitney - calculates U value ordinal (non-parametric) difference independent groups (unrelated)	Wilcoxon - calculates T value ordinal (non-parametric) difference repeated measures (related)	Spearman’s Rho - calculates rho value ordinal (non-parametric) correlation repeated measures (related)	Pearson’s r - calculates r value interval (parametric) correlation repeated measures (related)
Related T - calculates T value interval (parametric) difference repeated measures (related)	Unrelated T - calculates T value interval (parametric) difference independent groups (unrelated)	Chi Square - calculates X2 value nominal (non-parametric) difference/association independent groups - data in each cell (unrelated)	Sign test - calculates s value nominal (non-parametric) difference repeated measures/matched pairs (related)

What is the level of probability we normally use to interpret statistical tests? (1)

• what does this mean in terms of probability and chance? (1)

• when is a more stringent/demanding level used? (1)

P ≤ 0.05

→ 95% probability that result is not due to chance

In clinical trials:

P ≤ 0.01 (99% probability)

Difference between type I and type II error (1)

• effect of each (1)

Type I error = the null hypothesis is wrongly rejected (false positive) → so false alarm/believing there is an effect when there isn’t

Type II error = the null hypothesis is wrongly accepted (false negative) → so believing there is no effect when there is

How to reduce Type I (2) and Type II (2) error

Type I:

• set more stringent/more demanding significance level (e.g. use 0.01 rather than 0.05)

• replicate studies

Type II:

• set a less stringent/less demanding significance level (e.g. use 0.05 rather than 0.01)

• increase the sample size

Define test statistic

Define critical value

Define significance

Test statistic - calculated/observed value after using statistical test

Critical value - value that calculated value/test statistic must reach to show significance (selected from critical values table)

Significance - indicates research findings are sufficiently strong so researcher can reject null hypothesis and accept alternative hypothesis

Difference between one-tailed and two-tailed test

One-tailed = has directional hypothesis

Two-tailed = has non-directional hypothesis

Define ‘N’

Define ‘Df’

What is the importance of ‘R’ for determining significance?

N = number of participants

Df = degrees of freedom, number of values that are free to vary

Importance of R:

• tests with R in name → calculated value must be gReater than critical value = significance

• tests without R in name → calculated value must be less than critical value = significance

How to carry out the sign test to find significant difference (7)

• ensure you consider how to find the critical value from a table (3 aspects to remember)

1. organise results data into table

2. calculate the differences between DV of the two conditions

3. count the number of positive signs and negative signs (in the differences)

4. determine which sign is less frequent = the S value (calculated value)

5. find critical value from table, considering if test is one or two-tailed, level of significance (usually 0.05), and number of ppts

6. compare S value to critical value

7. if S value is less than critical value, then difference is significant = null hypothesis rejected, alternative accepted