scientific processes

aim

a statement of the study’s purpose:

• research should state its aime beforehand so it’s clear what the study intends to inv.

alternative hypotheses: one-tailed hypothesis

(directional) predicts a directional difference between your variables.

• e.g. higher, lower, faster slower

• ‘students w/ higher attendance will have significantly higher grades’

alternative hypotheses: when to use a one-tailed hypothesis?

when you have previous research findings which suggest which way the results will go.

alternative hypotheses: two-tailed hypothesis

(non-directional) predicts a difference but doesn’t state where the diff. lies.

• ‘there will be a significant difference in the grades of students w/ high attendance and students with low attendance.’

alternative hypothesis: when to use a two-tailed hypothesis?

when there is no or little previous research in the area or when results are mixed/inconclusive.

null hypothesis

what you assume is true during the study - any data collected will either back this assumption or it won’t:

• predicts there is no difference/relationship between your variables.

• if data doesn’t support null hypothesis → rejected and the alt. hypothesis (one or two-tailed) = accepted.

• ‘Tooth flossing has no effect on number of cavities.’

sampling: sample

a representative group of ppl from a target population.

sampling: population

all the ppl in a particular group e.g. a certain age, background.

sampling: random sampling

when every member of a target group has an equal chance of being selected for the sample - manually or by a computer.

sampling: 2 advantages of using random sampling?

• fair - everyone has a chance of being selected.

• sample = likely to be representative.

sampling: 2 weaknesses of using random sampling?

• not guaranteed to have a representative sample.

• researcher may end up w/ a biased sample as sample = too small.

sampling: volunteer sampling

when ppl actively volunteer to be in a study by responding to a request for pps advertised by r e.g. newspaper, notice board.

sampling: 2 advantages of using volunteer sampling?

• a large no. of ppl may respond.

• in-depth analysis + accurate results if larger sample.

sampling: one weakness of using volunteer sampling?

• not representative - only a sample of ppl who have responded.

sampling: opportunity sampling

when the r samples whoever is available + willing to be studied e.g. r’s working in universities use samples of students.

sampling: one advantage of using opportunity sampling?

• quick + practical

sampling: 2 weaknesses of using opportunity sampling?

• unlikely to be a representative sample.

• cannot generalise the findings.

sampling: stratified sampling

where important subgroups in the population e.g. diff age groups are identified + a proportionate no. of each is randomly obtained.

1. class of 20: 10 = 16 yrs old, 8 = 17 yrs old and 2 = 18 yrs old.

2. if you take a stratified sample of 10 students - the no. of 16, 17 and 18 year olds needs to be 50% of the whole class.

   • 5 16 yr olds.

   • 4 17 yr old.

   • 1 18 year old.

sampling: one advantage of using stratified sampling?

• fairly representative sample.

sampling: 2 weaknesses of using stratified sampling?

• time-consuming - all potential pps need to assessed + categories.

• some groups w/in a sample may not be represented if a small sample is used.

sampling: systematic sampling

where every nth name from a sampling frame (a record of all names in a pop) is taken.

• e.g. every 3rd name from a register or phone book.

sampling: 2 advantages of using systematic sampling?

• simple + effective way of generating a sample w/ a random element.

• pop. = more likely to be evenly sampled.

sampling: 2 weaknesses of using systematic sampling?

• subgroups might be missed.

• not necc. representative if pattern used for samples coincides w/a pattern in the pop.

sampling: what are the 5 sampling technqiues?

• random

• systematic

• stratified

• opportunity

• volunteer

pilot study and aim

a ‘feasibility’ study - a small-scale preliminary study conducted before any large-scale quantitative research in order to evaluate the potential for a future, full-scale project:

• fundamental stages of te research process.

• allow r’s to check methodology, standardise instructions + allow research to be conducted.

experimental design: independent groups

diff. pps in each group.

• one group of students use a new study method, another groups uses a traditional method.

experimental design: 2 advantages of independent groups?

• no order effects.

• fewer demand characteristics.

experimental design: 2 weaknesses of independent groups?

• pp variables - individual diffs.

• no. of pps - twice as many.

experimental design: repeated measures

all pps do all conditions - used to compare each condition to each other.

• pps tested on their ability to recall words while listening to diff. types of music.

experimental design: 2 advantages of repeated measures?

• pp variables (ind. diffs) = controlled e.g. age, intelligence.

• fewer pps needed (unlike independent groups).

experimental design: 2 weaknesses of repeated measures?

• order effects e.g. get better over time, fatigue → focus

• demand characteristics

experimental design: matched pairs

pps = paired based on similar characteristics - one member of each pair = randomly assigned to the experimental group the other to control group.

experimental design: 2 advantages of matched pairs?

• no order effects - diff. pps assigned + exposed to one condition.

• pp variables = controlled - matching e.g. IQ, age, gender that could affect outcome - controls for potential confounding variables (initial knowledge, motivation)

experimental design: 2 weaknesses of matched pairs?

• no. of pps - two groups of pps. (more than repeated measures)

• practicalities - time consuming, difficult to find ppl who match.

experimental design: what are the 3 experimental designs?

• repeated measures

• independent groups

• matched pairs

observational design: behavioural categories

categories defined by the researcher to observe during the experiment.

• e.g. aggression in children

observational design: event sampling

researcher records every event (if a behavioural cat.) when observed in the research.

• how many times a child initiates social interaction w/ peers.

observational design: time sampling

researcher records every behaviour w/in a certain time frame e.g. 10 mins.

• a child’s activities every 5 minutes during a free play period to understand their interests + engagement.

questionnaire construction: questionnaires

a form of self-report that can be written, face-to-face, phone or online.

questionnaire construction: open questions

qs that allow the pps to respond in any way and in as much detail as they like.

• gives detailed, qualitative info.

questionnaire construction: closed questions

qs that limit the answers that can be given.

• gives quantitative data = easier to analyse.

questionnaire construction: one advantage of using questionnaires?

• practical - you can collect a lot of data.

questionnaire construction: 3 weaknesses of using questionnaires?

• bad qs - leading qs or unclear

• biased samples

• self-report - lying, social desirability bias.

questionnaire construction: social desirability + its effect on validity?

when ppl try to show themselves in the best possible light - might not be completely truthful but give answers that are more socially acceptable.

• not a true rep. of someones opinions/thoughts/feelings → dec. validity.

interviews: what are interviews?

a form of self-report used to gather qualitative research which can be face-to-face or over the phone.

interviews: 2 advantages of using interviews?

• rich data

• pilot study

interviews: 2 weaknesses of using interviews?

• impractical - time-consuming (structured, lots of ppl), hard to rep., smaller sample sizes

• ethical concerns - sensitive topics (personal, emotional e.g. MH) → distress, embarrassment, uncomf.

interviews: structured interview

fixed set of qs that are the same for all pps.

interviews: 3 strengths of structured interviews?

• can be easily repeated/replicated

• requires less skill

• easier to analyse

interviews: 3 weaknesses of structured interviews?

• interview bias can still occur (int. unintentionally influences pps responses → accuracy e.g. tone., body lang)

• social desirability

• data collection = restricted by pre-determined qs

interviews: unstructured interview

interview starts w/ some general aims + qs and then lets interviewees answers guide subsequent qs.

interviews: 3 strengths of unstructured interviews?

• detailed info can be obtained

• high validity - good rapport (close harmonious relationship, understand each other’s feelings, good comm)

• deeper insight into thoughts + feelings

interviews: 3 weaknesses of unstructured interviews?

• interviewer bias

• requires training

• harder to analyse

interviews: explain how interviews can improve upon the data collected from analysis.

by asking follow up qs to gain greater insight into points collected:

• → more detailed info than pp would’ve been able to reveal.

• → further measure of rel.

• → interviewer gains rapport w/ pp - felt comf. rev. personal info f-f.

variables: independent variable

the variable directly manipulated by the researcher i.e. what is being changed that affects the DV.

variables:dependent variable

the variable you think will be affected by changes in the IV.

variables: operationalisation of variables

describes the process by which the variable is measured.

• allows other researchers to see exactly how you’re defining + measuring variables.

• taking a vague idea and defining it in a way it can be measured.

• e.g. aggression → counting punches in a game.

variables: extraneous variables

any variable other than the IV that could affect what you’re trying to measure, if not controlled.

• age, intelligence, mood, background knowledge

variables: how to control extraneous variables?

using random sampling - creates more equality between groups - spreads differences evenly.

variables: confounding variables

any variables that influence your DV other than the IV that we cannot be sure of the true source of changes to the DV.

• subset of extraneous variables that actually interfere w/ the results - harder to know whether the IV is truly causing the effect.

• e.g. diet in an exp inv if the exercise (IV) causes change in weight loss (DV)

variables co-variables

the variables being measured in a correlational analysis to inv. a relationship between them.

controls: what are the 4 ways to control variables?

• random allocation

• counterbalancing

• randomisation

• standardisation

controls: random allocation

everyone has an equal chance of doing either condition.

controls: counterbalancing

mixing up the order of the tasks - helps w/ order effects in repeated measures.

controls: randomisation

when materials are presented to the pps in a random order.controls:

controls: standardisation

everything should be as similar as possible for all pps.

demand characteristics

pps may have determined the aims of the study so act deliberately to please the researcher (or the opposite).

how to control for demand characteristics?

using counterbalancing or randomisation.

double blind

neither pps or researchers know which condition the pps are in.

• used in medical trials.

single blind

pps do not know what condition they’re in.

• used for exp + control groups.

ethics: ethical guidelines

dev. for psychologists to follow when designing studies so pps = protected.controls:

ethics: informed consent

pps should always give informed consent (under 16 by parents or guardian).

• deal w/: should be told aims + nature of study and the right to withdraw.

ethics: deception

if pps have been deceived - cannot give informed consent.

• dealt w/: sometimes researchers must withhold info about study as pps wouldn’t behave naturally if they knew what the aims were.

ethics: right to withdraw.

pps are allowed to withdraw from the research at any point.

ethics: confidentiality

none of the pps in the study should be identifiable from any reports that are produced.

• dealt w/: data must be confidential + anonymous.

ethics: protection from harm.

risk of harm to pps should be no greater than they would face in their normal lives.

ethics: why can a debrief be used as a method of dealing w/ all issues.

returns pps to the state they were in before the research.

• researchers must fully explain what the research inv. + what the results might show.

peer review: role of peer review in the scientific process?

involves all aspects of the psych. research being scrutinised by a small group of experts in a particular field.

• peers should be objective + unknown to the researcher(s).

peer review: 3 main aims of peer review?

• to allocate research funding - gov. run funding organisations (medical research council)

• to validate the quality + relevance of research (formulation of hypotheses, methodology chosen, statistical tests).

• to suggest amendments or improvements - work is not app. + should be withdrawn.

peer review: role of anonymity in peer review?

peer should remain anonymous to maintain an honest appraisal:

• a minority of reviewers may use their anonymity to criticise rival researchers who’ve crossed them in past.

• direct comp for grants/funding.

• some journals make names of reviewers public.

peer review: role of publication in peer review?

there is a tendency for editors to publish signif. ‘headline grabbing’ findings to inc. credibility + circulation of publication:

• could create a false impression of current state of psych if editors = selective.

• file drawer problem - negative results are not published (esp. meta-analysis = non-signific).

peer review: role of burying ground-breaking research in peer review?

reviewers = critical of research that contradicts their own + favourable to those who match theirs:

• established scientists → more likely to pub. research that ‘fits’ w/ current ops, rather than new + innovative that challenges.

• in turn, slows down rate of change.

implications of psych. research for economy: explain how the dev. of treatments for mental illnesses using psych research has economic imp.

absences at work cost economy estimated £15 billion per yr - a third caused by dep, anx + stress (the Telegraph, 2014):

• CBT/SSIRs + anti-anxiety drugs allowed ppls w/ mild MH disorders to return to work + access medical treatment.

implications of psych. research for economy: explain how psych. research into sleep beh + shift workers (Czeiser et al) has economic imp.

studied workers at factory whose shift patterns caused sleep + health probs:

• rec. rotating shifts every 21 days + changing shifts forward in time.

• working during night for a week, late afternoons for the second week + morning for the third, and returning to the night work.

• changes indicated inc. productivity + job satisfaction.

reliability: reliability

the overall consistency of a measure.

reliability: internal reliability

the extent to which a test is consistent w/in itself.

• e.g. whether the diff. qs in a questionnaire are measuring the same construct.

reliability: external reliability

the ability of the test to produce the same results each time it’s carried out.

• e.g. taking an IQ test one yr, taking the same test a year later → similar score → ext. reliability.

reliability: what are the 2 methods to assess reliability?

• split-half method

• test-retest method

• inter-rater observer reliability

reliability: explain how to use inter-rater observer to assess reliability.

when the consistency between measurements/obs/assess. of 2 or more researchers = assessed by asking each to categorise independently.

• data = compared + similar → reliable.

reliability: explain how the split-half method is used to assess reliability.

compares an ind. performance on two halves of a test.

• e.g. odd vs even qs and comparing scores on each half - similar results on both = reliable test.

reliability: explain how the test-retest method is used to assess reliability.

a person repeats a test a month or so after doing it the first time:

• as time has passed, pps shouldn’t remember their answers.

types of validity: (internal) validity

accuracy - whether something is true/legitimate.

e.g. study to test whether sleep affects memory but pps allowed to drink caffeine before bed.

types of validity: internal validity.

the study measures/examines what it claimed to measure/examine.

e.g. Loftus & Palmer - use of leading qs used the same critical q (changing the verb only) across 5 conditions of IV.

types of validity: external validity

the extent the results of the study can be generalised to others.

• ex = ecological validity

types of validity: ecological validity

a type of external validity - generalisable to real life settings - generalising findings from one setting to another.

e.g. Rutter et al.’s Romanian orphans study - inv. real impact of dep. due to being raised in an inst.

types of validity: concurrent validity

results from a new test can be compared to a previously well-established test.

• a way to assess validity.

• e.g. pps given IQ test - scores compared to scores on another - if correlate/similar = concurrent validity.

types of validity: population validity

whether you can reasonably generalise findings from your sample to a larger group of ppl.

e.g. social anxiety using 15 yr old girls - can’t generalise to adults males.

types of validity: temporal validity

assesses to what degree research findings remain over time.

e.g. Asch’s study = ‘child of its time’ - pop. who just came out of WWII - same lvls of conf. = unlikely today.

types of validity: face validity

the extent to which a test appears to measure what is intended to measure.

• a way to assess validity.

• e.g. does this digit-span test measure the capacity of ST memory?

types of validity: what are the 5 types of validity?

• face

• concurrent

• ecological

• temporal

• population