RM 2

• Quantitive data (numerical)-

  • closed questions
  • lab experiments
  • structured observation and interview
  • S- easily collected and analysed
  • S- able to see patterns and trends
  • S- less likely to be bias
  • W- lacks depth and detail
  • W- doesn’t tell much about real-life.
• Qualitative data (non-numerical)-

  • open questions
  • unstructured interview and observations
  • S- detailed insight
  • S- higher external validity- more meaningful
  • W- hard to replicate
  • W- time consuming
  • W- subject to bias
• Primary data (1st hand by researcher)-

  • specific to the investigation
  • experiment
  • observation
  • questionnaire
  • S- internal validity
  • S- recent/contemporary
  • S- fits the purpose
  • W- time consuming
  • W- expensive
  • W- potential bias
• Secondary data (existing)-

  • meta-analysis
  • journal articles/books/websites
  • S- cheap and easy
  • S- less bias potential
  • S- greater external validity
  • W- uncertain research quality/validity
  • W- possibly outdated
  • W- not fit researchers purpose
• Role of peer review-

  • small groups of experts scrutinise the research before it’s released
  • they’re objective and unknown to the researcher/reviewer to keep it honest and reliable
  • to ensure accuracy of findings
  • to allocate funding for future research
  • ensure quality and relevance of research
  • prevents irrelevant findings being published
• Process of peer review-

  • works checked by other psychologists/independent experts, before deciding if the work should be published.
  • can be anonymous
  • recommended changes/revisions or recommended for publishing/rejection
  • check work for validity, originality and significance
  • editor of journal makes the final decision on rejection/publication
• Peer review evaluation-

  • S- anonymity prevents bias, makes it reliable and honest.
  • W- bias in publication, due to companies false listing journals for gain (keep the truth hidden)
  • W- keeping the status-quo (not publishing ground breaking theories) lacks progress
• Correlations-

  • the strength and direction of an association between 2 or more co-variables.
  • co-variables are variables being measured for an association using a correlation.
  • weak or strong
  • positive or negative or moderate
  • positive = increase left to right
  • negative = increase right to left
  • skewed = scattered/no pattern
• Correlation studies-

  • can’t reliably establish cause and effect
  • predicts a relationship between variables, not a difference.
  • lacks control over (intervening) variables

  VS

  Experiments-

  • have (in)dependent variables
  • able to observe cause and effect
  • controls and environments can be set up.
• Correlations

  Evaluations-

  • S- useful for primary research, to identify a link for future research.
  • S- useful for sensitive or unethical research (no deliberate manipulation of variables)
  • W- doesn't identify a causal relationship
  • W- problem of a 3rd variable
• RM- content analysis

  • Indirect observational study of behaviour through media (diary, speech, tv, texts)
  • Used if its unethical to study 1st hand
  • Quantitive information of themes occurrence in a tally shown in tables, graphs, bar charts
  • Thematic analysis- qualitative approach of gathering themes - broader/descriptive
  • W- subjective to interpretations + indirect
  • S- easily analysed + ecological validity + ethical + qualitative/quantitative
• RM- Case Studies-

  • individuals + unique + longitudinal + qualitative
  • interviews (unstructured), observations (c/overt + unstructured), closed questionnaires
  • S- in-depth data = W- lacks generalisability
  • S- furthered understandings of ‘normal’ behaviours through comparison
  • W- subject to interpretation/bias + retrospection = impacted validity
• RM- Reliability

  • test-retest- same results after being retested - scores correlated to test relatability - needing +0.8
  • Inter-observers- agreement of 2/+, can have a pilot study, operationalised categories, same things viewed, answered individually
  • Improvements-
  • questionnaires= closed questions, +0.8 retest
  • Interviews= structured + closed questions, training, test-retest, inter-interviewers
  • Experiments= replication, pilot study, standardised procedure
  • Observations= recorded study, simple + separate categories
• RM- Validity

  • internal= within the study, effect of the IV on the DV - effected by extraneous/confounding variables - identify before the experiment - lab + double blind experiments
  • External= ecological- real world, population- groups - effected by samples, setting, task - improve realism with tasks/situations
  • Temporal= how theories change over time (Asch 1960’s had the fear of communism - not elsewhere)
  • Face= simple test, from an expert to see if it’ll measure what’s wanted
  • Concurrent= new compared to a similar established study - valid with a +0.8 correlation
• RM- levels of measurement

  • Nominal= categories, shown in a bar chart, mode (central tendency), BASIC
  • Ordinal= ranked, continuous data, shown in a histogram/line graph, subjective gaps, CT=median, dispersion=range
  • Interval= mathematical intervals, CT=mean, dispersion=standard deviation, CAN BE MADE INTO ANY LOWER LEVEL
• Statistical tests-

  • Determines significance of results + IV/chance on DV + accept/reject null hypothesis
  • Need to: identify levels of data, repeated measures/independent groups, test for difference/correlation
  • P </= 0.05 (level of significance 0.01 if risk on life)
  • Critical value= 1/2 tailed, number of participants, significance (0.05/5% unless stated)
• Critical value table-

  Carrots Should Come Mashed With Swede Under Roast Potatoes

  
• Sign test-

  1-Find the difference between the 2 sets of numbers, by seeing if it increases or decreases.

  2- Add up all the ‘+’ and ‘-‘ signs, not ‘=‘

  3- The least frequent sign becomes the ‘S’ value.

  4- Find the number of significant participants

  5- Use the critical value table and use the correct column for a non/directional hypothesis (1/2 tailed).

  6- Then use the 0.05 column of P’s to have the CV

  7- If the critical value is equal or less than the ‘S’ value, then it is significant. If the ‘S’ value is higher than the critical value = no significance
• Experimental method-

  • Aim- what, why + psychological theory developed
  • Hypothesis- testable prediction, states variables relationship - null= no effect, alternate- directional= what difference is, non-directional= difference
  • Operationalised= defined variables + how they’re measured
  • Independent variables- different conditions + controlled
• Research issues-

  Extraneous variables- unwanted, found before the experiment, limited impact

  Confounding variables- systematic + independent, makes it hard to see the DV change + effect

  Demand characteristics- clues of the experiment

  Investigator effect- un/conscious influence on DV

  Randomisation- chance being the basis of experiments design

  Standardisation- all the same for each condition
• Experimental design-

  Independent groups- different levels of experiment, all separate to compare results - W- participant variables, economically wasteful

  Repeated measures- same P’s, tested both scores - W- bias from doing the experiment x2 = S- counterbalancing (AB/BA)

  Matched pairs- P’s paired for wanted variables, different conditions (controlled confounding/P’s variables) - S- no demand characteristics - W- economically consuming + never a full match
• Types of experiments-

  Lab-

  • Controlled + artificial
  • S- extraneous variable control - DV only affected by the IV = internal validity
  • S- easily replicated
  • W- artificial = lacks ecological + mundane validity
  • W- P’s know they’re watched
• Types of experiments-

  Field-

  • natural setting, IV controlled, identify casual relationships
  • S- mundane realism + authentic actions
  • CPS- external validity - unaware of the experiment
  • CPW- ethical issues - no consent + privacy
  • W- not controlled environment + hard to see cause + effect + hard to replicate
• Types of experiments-

  Natural-

  • no direct control of the IV = natural change, can be in a lab
  • S- ecological validity
  • W- cant be randomly allocated, hard to see IV cause + effect
• Types of experiments-

  Quasi-

  • IV from real differences between people, not randomly allocated, IV cant change, DV can be natural or created
  • S- controlled, replicated
  • W- possible confounding variables, IV cant be seen to have an impact
• Sampling-

  Random -names = numbers, lottery method to

  choose sample - s- internal validity - W- time consuming, may refuse

  Systematic- mathematically chosen, nth person, alphabetical order - S- no bias - W- time consuming, may refuse

  Stratified- groups of people, some selected fro sample = broad

  representation - s- broad representation - W- personal differences

  Opportunity- convenience - s- time/ money effective -

  W- unrepresentative, possible controlled selection

  Volunteer-self-selecting - s- simple, engaged - W- attract a certain Profile
• Ethical issues-

  • British Psychological Society code of conduct = ethical guidelines, forced by ethic committees
  • Informed consent- letter, parental consent (<18), presumptive= similar group, prior= multiple before, retrospective= at debrief
  • Deception- debrief, allowed to withdraw, what data‘s used, counselling, reassured they’re normal
  • Privacy- no details kept/published, numbers/initials
• Pilot studies-

  • small scale prior to the experiment = potential issues
  • single blind- P’s don’t know the aim = control + debriefed
  • double blind- no-one knows the aim = no expectations
  • control- not changed = benchmark
• Features of science -

  Replicability -

  • Finding out if a method is valid and if its results are reliable
  • Repeated over different contexts and circumstances
  • Checking to see if faults are accurate or anomaly
• Features of science -

  Objectivity -

  • no observer input/bias = objective
  • ensured by removing anything that’s subjective to interpretation/bias
  • standardised instructions + fully operationalised variables
  • subjective = bias of the researcher on the P’s = not independent research
• Features of science -

  Falsification -

  • scientific principle wasn’t always true even if repeated/successful — just shows that the theory was strong
  • ‘all swans are white’ - only falsified by finding a black swan, not by many white swans
  • reasoning why you cant use ‘prove’
• Features of science -

  Paradigms -

  • scientific disciplines are seperate from non-scientific — scientific share a set of assumptions + methods
  • social sciences = ‘pre-science’
  • natural science = have core principles
  • psychology = ’pre-science’ - conflicting research
  • to be a science there must be a scientific revolution — contradiction = paradigm shift
• Features of science -

  Theory construction + hypothesis testing -

  • theory = general laws/principles that explain different behaviours/experiences
  • theory construct = empirical evidence/methods
  • theories need to have a hypothesis that can be tested systematically and objectively
  • deduction = creation of a new hypothesis from existing theories