Research methods Flashcards
What does HMC SQUID FORCE stand for?
H - Hypothesis testing
M - Manipulation of variables
C - Control of variables
S - Standardisation
QU - Quantifiable measurement
I - Induction
D - Deduction
F - Falsification
O - Objectivity
R - Replicability
C E - Cause + Effect
What is Hypothesis testing?
procedure that uses sample data to evaluate the credibility of a hypothesis about a population
What is the Manipulation of variables?
we manip./change IV ∴ can infer cause + effect
What is the Control of variables?
- Making sure EV don’t affect outcome of study
- controls = anything we do to try + stop these affecting outcome (DV)- ⬆se our confidence in cause + effect
What is Standardisation?
Ensures RS’s carried out in consistent way:
- experience of ps = same every time - standardising procedure, materials + instruct.
- Helps control Vs + make RS replicable
What is Quantifiable measurement?
- Ensuring numerical data’s gathered - allows to statistical analysis + compare results
- Stat. tests used to decide if res = sig. (cos of IV)
What is Induction?
- General conclusion, principle/explanation derived by reasoning from particular instances/obs
What is Deduction?
- process of reaching specific conclusion, b4 studies carried out, make generalised Hp ➡ then test
What is Falsification?
- Being able to prove something wrong (can Ffy. something if 1 E.G that goes against belief found)
- Must ensure results we collect have pt. to prove theory wrong.
- Unscientific rs. may not have pt. to be ffied
What is Objectivity?
- all sources of bias = minimised + personal/subjective ideas = eliminated
What is Replicability?
What is Cause + Effect?
- Cause states why something happens.
- An effect states a result/outcome.
- At times, a single cause ➡ several effects.
What’s peer review?
- Process that occurs b4 a study’s published = check quality + validity of RS, + ensure RS contributes to its field
Research cycle?
Theory ➡Hp(predictions) ➡ experiement ➡analyse results(draw conclusion)➡
Sections of journal/report: Abstract
- 1st section in report but wrtiten @ end
- Includes summary of: aims, Hyp, method, results + concl, + ∴ provides overview of entire report
Sections of journal/article: intro
- Justifies reasons for writing ab ur topic.
- Intros topic to reader, overview of previous RS on topic provided, key terms defined + own hyp identified
Sections of journal/article: method
4 subsections: - exact procedure
1 - design: IV + DV, EV, exp. design
2 - particip.: who, how many, how u got them
3 - materials - specifically what u used
4 - exact procedure- what u did
Sections of journal/article: Results
- calc mean, mode, range + median(measures of central tendency) and SD
- graphs + tables
- stat. analysis - Will Coxon - Chi-squared
- verbal explanation of findings
Sections of journal/article: Discussion
- ref. bk. to rationale(aims) of study + expl what progress RS made in its field
- compare what’s found in results sect. to what other ppl found in intro + pt. offer expl - Y u found what u did
- limits in method(criticism):strengths + weaknesses of study + ways to overcome in future
Sections of journal/article: References
- ref every psch. talked ab
- end of report
- alphabetical order
- Author name / initials, date of publication,title of article/ journal volume
Sections of journal/article: Appendix
- Raw data, mathem. calc., standardised instruct, lists of words, questionnaire used in RS, detailed description of an apparatus used in RS, stat test workings out
IV
V we manipulate (cause)
DV
we meas. - see if IV has effect on DV/effect
Conditions
group of particip. exposed to IVs
EVs:
- could pt. have effect on DV other than IV
Participant Vs:
- any individual diff. between ps that may affect DV
Situational Vs:
- could pt. lower internal validity = Rser = less confid
- IVs affected Dv + means what they wanted
Control of EV + CV: Randomisation -
- use of chance in order to control for effects of bias when designing materials deciding order of conditions - way to control investigator effects
Operationalising variables
- putting abstract concepts ➡ into concrete measurable concept (vs into a form they can be tested)
= being more specific
CV(confunding)
- found after study + these can be found to have affected DV - change systematically w IV ∴ can’t be sure if any obs changed
- lower internal validity of study = Rser knows DV isn’t direct result of IV
Control of EV + CV: Standardisation
- Using some formalised procedures + instructions for all ps in Rs study - w/in a study as far as possible all ps should be subject to same environ., info + experi.
Control of EV + CV: Reduce demand characteristics
- hide aim (using deception/distractor Qs)
Describe exactly Harvard referencing of: Books
- Surname, Initial (of 1st name).(date).
- Title of book in Italics,
- Where it was published: Publisher who published it.
Describe exactly Harvard referencing of: Journal articles
- Surname, Initial (of 1st name).(date).
- Title of book in NOT in Italics .
- Title of journal in Italics. volume number. pg numbers.
Similarities of Harvard ref. in books and journal articles:
- comma after surname
- brackets for date
- start with surname
Differences of Harvard ref. in books and journal articles:
- title of JA bot in italics but B is
- punctuation is diff.
- books have place but articles don’t
- JAs always have pg #, Bs don’t
Control of EV + CV: having control group
don’t take part in IV - acts as a baseline
what’s the RS aim?
Statement of which RSer intends to find out in RS study, always starts w/ ‘to investigate…’
what’s the RS Q?
Q Rser is trying to answer
Hypothesis:
- Precise + testable statement/ prediction ab assumed relationship betw Vs
what do we use when we use experimental methods?
use experimental Hp
what do we use when we use non-experimental methods?
non exp. methods (obs, self report + correlation) - alternate Hp
Null Hyp:
- says there’ll be no diff betw. conditions
- Rsers should aim to support null Hyp. = avoid Rs bias
Alternate Hyp:
states that there’s relationship betw the 2Vs being studied (IV has effect on other)
One-tailed (directional) Hyp:
- when u say what direction o effect will acc be
- includes words like more/less, higher/lower, faster/slower
- Rsers tend to use directional Hyp. when findings of prev studies sugg particular outcome
two-tailed (non - directional) Hyp:
- when u say there’ll be a difference/relationship but don’t say what that diff. will be
- Rsers use when no prev Rs/prev. Rs that’s been conducted in Area = contradictory
Alternate Hyp - One-tailed (directional) : writing frame
Ps who [IV condition 1] will…sig.ly more/less [DV] that Ps who [IV condition 2]
Alternate Hyp - two-tailed (non - directional) : writing frame
there will be a sig diff in the [DV] of the Ps [IV condition 1] compared to Ps [IV condition 2]
Null Hyp : writing frame
there will be NO sig diff in the [DV] of the Ps [IV condition 1] compared to Ps [IV condition 2]. Any diff will be due to chance
Laboratory experiment:
- RSer has strict control over Vs + uses standardised procedures in controlled environm.
- Cause + effect can be established as EVs are minimised
- Laboratory exper. tend to take place in artificial environments
- RSer manipulates IV
Laboratory experiment: Strengths
- ⬆internal validity – ⬆ controls so only IV should affect DV
- ⬆ scientific – easy to repeat (standardised procedures)
- ⬆able to assume cause and effect
Laboratory experiment: examples -
- Loftus + Palmer
- Bandura
- Moray
- Simon + Chabris
- Blakemore
- Cooper
Laboratory experiment: weaknesses
- ⬇ ecological validity – hard to generalise findings of lab experiments to everyday life
- ⬆ed demand characteristics – pps alter natural behaviour to ‘impress’ experimenter
Field experiments:
- Exp. takes place in subject’s own environ, but RSer still manipulates IV
- This type of exp moves out of artificial lab + sets itself in ⬆ly natural environment
- investigator creates situation of interest + then records people’s reactions + behaviours.
Field experiments: Strengths
- ⬆es ecological validity of findings in comparison w/ lab exp.
- Ps = less likely to display demand charact. or evaluation apprehension.
- Experimenter maintains control of the IV, + ∴ cause + effect relationships are still possible.
Field experiment: examples -
- Chaney
- Piliavin
Field experiments: weaknesses
- Changes in IV may be subtle + go unnoticed by Ps.
- Reactions of Ps may go unnoticed by experimenter.
- V little control over EV
- Ethical implications as Ps not aware of their involvement ∴can’t give consent.
Quasi-experiments:
- IV is already naturally occurring – RSer just records effect on DV
- Experiment may take place in a lab or in natural environment
- RSer examines naturally occurring IV
Quasi experiments: Strengths
-⬆er ecol. validity
- If subjects = unaware of being studied will be v little bias from demand characteristics.
- Allows RSer to investigate As that would otherwise be unavailable to them.
-⬆ed validity of findings due to lack of experimenter manipulation.
Quasi experiment: examples -
- Sperry
- Casey
- Baren - Cohen
- Levine
- Maguire
Quasi experiments: weaknesses
- Difficult to infer cause + effect due to lack of control offer EVs + no manipulation of the IV.
- Impossible to replicate exactly
- May be subject to bias if ps know they’e being studied.
-Ethical issues of consent, deception + invasion of privacy.
Difficult to infer C + E = lack of control over EVs + no manipulation of IV
Impossible to replicate exactly
May be subject to bias if ps know they’re being studied
Ethical issues of consent, deception + invasion of privacy.
Independent measures design:
- AKA between-groups
- diff. Ps used in each condition of IV
- ∴ each condition of experiment includes a different group of Ps
Independent measures design: Strengths
- Less chance of demand characteristics – they only take part in one condition which it makes more difficult to work out the aim of the study and therefore less likely to change behaviour to suit what they think the experimenter wants
No order effects – participants do not get better (with practice) or worse (with boredom) in the second condition as in an independent measures design participants do not take part in a second condition
Independent measures design: weaknesses
- When using diff Ps in each
group, it’s impossible to control individual
differences (an EV) which
can = inaccurate results. - A Ler sample of Ps is req for independent measures = can be less time efficient.
Repeated measures design:
- AKA within groups, or within-subjects design.
- same Ps take part in each condition of IV ∴ each condition of experiment includes same group of Ps
Repeated measures design: Strengths
- It’s unlikely results will be distorted by individual differences as each condition
will see these differences ∴ allowing
more accurate results. - sample required may be smaller as
every participant can repeat conditions.
Repeated measures design: Weaknesses
- Ps might guess aim of task as they repeat experiment in different conditions, can = in demand characteristics.
- Order + fatigue effects can = an
inaccurate reflection of behaviours
=ed by varying conditions.
Target population:
total group of individuals from which the sample might be drawn
Sample:
- Sampling = process of selecting a representative group from population under study
- the group of people who take part in the investigation.
Difference between population + sample?
- pop. = entire group U want to draw conclusions about
- sample = specific group that you will collect data from
- size of sample always < total size of pop
Random sampling:
- Type of probability sampling where every1 in entire target pop. has an =al chance of being selected.
- involves identifying every1 in target pop + then selecting # of ps U need in a way that gives every1 in pop an =ual chance of being picked
Strengths of random sampling:
- least chance of bias(every1 in entire target pop. has an =al chance of being selected)
- most likely to be able to generalise
Weaknesses of random sampling:
- unlikely to be to be repping target pop. w/ small samples - unlikely to be able to generalise appropriately to every1
- likely to need L sample - ppl chosen may not want to take part ∴final sample = biased + not truly random
What does generalisability mean?
Extent to which we can apply the findings of our research to the target population we are interested in.
What does population validity mean?
The extent to which the results can be generalised to groups of people other than the sample of Ps used.
Bias:
is the tendency to make decisions or to take action in an unknowingly irrational way
Self-selected sample/volunteer:
consists of participants becoming part of a study because they volunteer in response to an advert.
Strengths of self-selecting/ volunteering:
- can target specific groups
- RSers get informed consent from beginning bcos P’s have signed up freely
- if issue = socially sensitive ppl can select selves knowing they might be upset by nature of study = more ethical
Weaknesses of self-selecting/ volunteering:
- likely to = biased - certain type of person likely to volunteer
- less able to generalise to rest of pop. = likely to be biased
Opportunity sampling:
- sampling technique most used by psych. students
- consists of taking sample from ppl who are readily available @ time, study’s carried out + fit the criteria you are looking for
Strengths of Opportunity sampling:
- Easiest sampling type for RSer = they can jus use whoever is there
Weaknesses of Opportunity sampling:
- most chance of sample being biased = unlikely to rep target pop.
- RSer may show selection bias + ask certain types of ppl to take part
- least likely to be able to generalise to rest of pop. = highly unlikely to rep them
Snowball sampling:
- where you obtain sample by identifying 1 or a few ps from target pop + they ask ppl they know to take part.
- Useful if you’re investigating hard to reach groups
Weaknesses of Snowball sampling:
- No way of knowing whether sample’s representitive of pop.
- suffers from gender,culture + age bias
Strengths of Snowball sampling:
- Possible to include members of groups where no lists or identifiable clusters even exist + who may not be easy to access
- especially useful for p’s who don’t want to be found out
Stratified sampling:
Classifying population into categories + then choosing a sample which consists of ps from each category in same proportions as they are in pop.
The British Psychological Society:
Representative body for psychology
British psychological society - 4 key principles:
RESPECT
COMPETENCE
RESPONSIBILITY
INTEGRITY
Repect:
- Valuing dignity + worth of all individuals
- Includes awareness of how psychologists may influence ppl + appear to have authority, + on people’s rights to privacy + self-determination
Respect: What is privacy and confidentiality?
Any data must be kept confidential with no identifiable information on it
Respect: how would you deal w/ privacy and confidentiality?
- Shouldn’t use p’s names in published work/allow them to be IDed in any way
E.G. in case study use diff. name/initial + avoid publishing details of address, skwl, fam, photos etc. - selection from L amounts of data may = observer bias
- findings from 1 individual can’t be generalised to others
Respect: What is Informed consent?
- Give ps every bit of information about the study so they can decide whether they want to take part of not
- If ps = children (U16) would need to get informed consent from parents
- Obs in public = don’t require consent
Respect: how would you carry out Informed consent?
- Give ps every bit of info
- If not possible = “lack of informed consent”
Respect: What is the Right to withdraw?
Pps have the right to withdraw themselves and their data at any point during or after the research has been carried out
Respect: how would you carry out Right to withdraw?
- Give them right to withdraw + remind them of it during study + @ end
- they should be told they can w/draw themselves +/or their data @ any time
Competence:
Valuing continuing development as psychologists and the maintenance of high standards at work
Enhancements of knowledge, skill, training, education and experience
Responsibility:
Important that psychologists protect their Ps + ensure that they’re full aware of the procedures that they’re taking part in
Responsibility: what is the protection of participants from harm?
Everything should be done to try + protect Ps from harm (both psych. + physical)
Responsibility: how would you ensure the protection of participants from harm?
- Should be careful when RSing sensitive topics
- procedures shouldn’t be stressful towards the P’s
Responsibility: What is debriefing?
- when U give Ps full details of what study was ab, what U intended to find out what you’ll do w/ results, if they’re happy to be a part = then keep their results
- should ensure ps leave the experiment same state as they began
Responsibility: Is debriefing an issue?
NO but is a way of dealing with ethical issues set out
What is integrity?
- Psychologists should maintain high standards regarding honesty, accuracy, clarity + fairness + should avoid deception where possible
Deception:
When U gives Ps false info + tell them Ur study is ab something that it’s not
How would you deal with deception?
Can’t acc deal w/ it once you’ve lied but are things you can do to make it better:
- debriefing
- Right to w/draw
Ethical considerations @ start:
- Gain informed consent from Ps
- give them right to w/draw @ any point
- Keep results confidential
Ethical considerations throughout:
- don’t lie to them
- remind them of their right to w/draw
- maintain confidentiality
- protect ps from harm
Ethical considerations after the study:
- debrief fully
- give them right to w/draw their data
- maintain confidentiality
How would you write a consent form? (comes @ start of study)
- Should tell ps: what study’s about - aims/purpose, what they’ll be expected to do
a. how long it will take
b. what the results being used for + that their info will remain confidential
c. That they have right t w/draw @ any point
d. that they’ll be protected from harm
e. that they’ll be debriefed @ end
How would you give standardised instructions (exams Qs)?
- instructions that get read out @ beginning of procedure
- It’s important that they’re standardised as all grps need to hear same info
- Also important = other Rsers may want to replicate Ur work + therefore need to use same instructions
- instructions should be written verbatim which = they can be read out to other P’s
How would you write a debrief (exams Qs)?
a. What acc aim was - if they weren’t told @ the beginning
b. what ALL the P’s were doing in the study - if it was independent groups or matched pairs (need to know what others were doing as well as selves)
c. That they still have right to w/draw
d. that their results remain confidential
e. that they have chance to speak to trained counsellor if they feel if they feel they’ve been emotionally/physically upset by procedure
What does RUMBA stand for?
Read the Q.
Underline key psychological words.
Marks analysed.
Box command words.
Annotate.
What would say if asked why psych carry out unethical RS, or adv of unethical RS (exam Qs)?
a. deception can reduce demand charact. = increasing validity
b. some -ve topics = difficult to study w/out possible distress to ps but understanding -ve behaviour is important
c. If guidelines have been broken then this could be as minimal as possible + debriefing etc should be used
VALIDITY:
Whether test does what it set out to do/measures what it set out to measure
INTERNAL VALIDITY:
refers to the degree of confidence that the causal relationship being tested is trustworthy and not influenced by other factors or variables.
FACE VALIDITY:
is about whether a test appears to measure what it’s supposed to measure. This type of validity is concerned with whether a measure seems relevant and appropriate for what it’s assessing only on the surface
CONSTRUCT VALIDITY:
Asks whether a measure successfully measures the concept it is supposed to (e.g. does a questionnaire measure IQ, or something related but crucially different?)
CONCURRENT VALIDITY:
involves comparing a new test with an existing test (of the same nature) to see if they produce similar results. If both tests produce similar results, then the new test is said to have concurrent validity.
CRITERION VALIDITY:
refers to the extent to which the results and conclusions are valid compared with other measures.
EXTERNAL VALIDITY:
extent to which the results of a study can be generalized to other settings (ecological validity), other people (population validity) and over time (historical validity).
POPULATION VALIDITY:
extent to which the sample can be generalised to similar and wider populations. This type of validity is important because without it the research becomes low in usefulness.
ECOLOGICAL VALIDITY:
is a measure of how test performance predicts behaviors in real-world settings.
Improving validity: Questionnaires -
- Many contain lie scale w/in Qs in order to assess constituency of respondent’s responses + control effects of social desirability bias = will be further enhanced if Q-aire is anonymous
Improving validity: Experiments -
- by having control group so that the RSer can compare effect of IV
- Procedures can also be standardised = will minimise impact of investigator + particip. effects (demand charact.) - use of controls is effective also
Improving validity: Observations -
- validity can be improved by ensuring behavioural categories aren’t overlapping or too broad
- Generally speaking obs. are high in ecol. validity especially if the obs. is covert
Reliability:
The extent to which findings from a study can be repeated
Findings must be consistent
Internal reliability:
Consistency of a measurement within the study – is the study standardised?
Experiments are often high in internal reliability as they are highly controlled
Field experiments tend to lack internal reliability due to lack of control
External reliability:
If the measurement was repeated it should get similar results (replicating)
E.g take an IQ test in January and then again in June and the scores should be the same
Inter-rater reliability
This is where 2 observers who are observing the same event in the same way and have a high level of agreement (usually 80%)
Test-retest
Once a researcher has done a test, if they do it again at a later date with the same pps it should have similar results
Split-half method
The researcher give the p’s 2 sets of questions all measuring the same thing – they answer Q 1-10 and get a score and then answer Q 11-20 to get a score and then compare the results.
