Research Methods year 2 Flashcards
(37 cards)
Content analysis + Coding
- Turns qualitative data into quantitative.
- Indirect study of peoples behaviour using communications eg speech, text, email = qualitative
- First stage = look through text and understand what it is so potential codes can be made
- Second stage = coding > data sets could be large so should be categorised into codes like themes, phrases, words. EG of code: no. of derogatory terms for mentally ill like mad.
- Read the text again and count the amount of times the codes appear = quantitative dat
Evaluation of content + thematic analysis
- Ethics aren’t usually an issue because the material is usually in the public domain > suggests because resources are already public, there is already consent etc
- Quantitative data can be a problem due to lack of understanding which means the individual experience is unacknowledged.
- Lack of objectivity when compared to thematic analysis. because research looks at issue indirectly + outside of the context it happened in. > researcher may attribute opinions which were not intended.
Thematic analysis (themes)
- Leaves qualitative data as qualitative but in less broad categories
- Indirect study of people’s behaviour using communications eg speech, text, transcripts. > qualitative
- First, identify themes from the source. (theme= ideas which are frequent + more descriptive than codes)
- EG of theme: in newspapers, mentally ill may be seen as “a threat to society”
- Could then be put into broader categories like stereotyping or treatment
- Direct quotes can be used to illustrate each theme > qualitative
Levels of measurement
- Factor which helps choose stat test
- Nominal
- Ordinal
- Interval
Nominal data
- Least amount of detail + in named categories + is discrete (only 1 item appears in 1 cat) > most basic form of data. Does not look at individual ppt, groups them together (boys or girls, not ppt 1,2,3)
- EG: no. of boys + girls who did and did not conform in a line test
Ordinal data
- Ordered data, involves rating and use of scales > makes it lack precision and subjective because it is based on opinions and not fact based.
- EG: rating of psychology from 1-10
Interval data
- Based on true numerical values, most amount of detail making it specific and precise. (can measure if something is exactly half of something unlike ordinal)
- EG: different in heart rate
how to choose an Inferential statistical test
- Test used to make inference about data.
- Is it a test of difference or test of association (correlation).
- Then have to check to see if it is a unrelated design (independent groups) or related design (repeated measures)
- Then choose whether it is suited to nominal, ordinal or interval data.
Inferential stats: Nominal
- Test of difference + unrelated: Chi-squared (X^2)
- Test of difference + related: Sign test (S)
- Test of association: Chi-Squared (X^2) association
Inferential stats: Ordinal
- Test of difference + unrelated: Mann whitney (U)
- Test of difference + related: Wilcoxon (T)
- Test of association: Spearmans rho (Rho: Rs)
Inferential stats: Interval
- Test of difference + unrelated: Unrelated t-test (t)
- Test of difference + related: Related t-test
- Test of association: Pearson’s r (r)
Features of science
- Objectivity and empirical method
- Theory and hypothesis testing
- Replicability
- Falsifiability
- Paradigm shift
Objectivity and empirical method
DEFINITION:
-This is where the researcher does not impose their own beliefs or ideas when testing theories
-The empirical method is where the data is collected through direct experiments + observations > not just based on argument or belief
EXAMPLE IN PSYCH:
-Behaviourist approach shows this as the conclusion that phobias are acquired through classical conditioning and maintained by operant conditioning are shown in a lab experiment of little Albert.
-Cognitive theorists show objectivity + empirical methods when establishing types of LTM as brain scans were used to observe brain activity when using different LTM
EXAMPLES ITS NOT IN PSYCH:
-Psychodynamic approach shows this as the activity Freud talks about is in the unconscious + Freud’s argument comes from case studies + belief
-Humanistic approach shows this as concepts such as self-actualisation and self-esteem cannot be measured.
Theory + hypothesis testing
DEFINITION:
-Theory > set of laws which explain behaviours. Hypothesis > prediction indicating what may happen in experiment + must be operationalised (testable)
Hyp has to be tested to see if evidence supports or refutes it. > if refuted it has to go through deduction (new hyp)
EXAMPLE IN PSYCH:
-Cognitive theorists in memory like Baddeley hypothesised that different coding was relevant to different memory stores > tested this in coding research with acoustic + semantic similar words > made his theory that STM=Acoustic and LTM=Semantic
-Behaviourists + little Albert
EXAMPLE IT IS NOT IN PSYCH:
-Humanism > you cannot test level of freewill or self esteem
-Psychodynamic > cannot scientifically test psyche or unconscious
Paradigm + Paradigm shifts
DEFINITION:
-Paradigm > set of shared assumptions + beliefs in a scientific discipline
-Paradigm shift > when an accepted paradigm is challenged with valuable evidence which is too great to ignore > there is a shift and scientific revolution
EVIDENCE IN PSYCH:
-Psychologists agree psychology is study of mine + behaviour =paradigm + broad agreement
-Has had paradigm shifts from Wundt to cognitive neuroscience
EVIDENCE NOT IN PSYCH:
-Lack set of shared assumptions > psych is more of a pre-science, have internal disagreements + conflicting approaches unlike biology or physics.
-Level of explanations rather than a set shared one.
Falsifiability
DEFINITION:
-idea that a theory cannot be seen as scientific without admitting to the possibility of being untrue or disproved
EXAMPLES IN PSYCH:
-Behaviourism > can be falsified as has testable methods. EG, in attachment ideas of learning through C/C has be refuted by Bowlby + Harlow+Harlow
-Biological > falsifiable because other impacts may be present. EG in biological exp of OCD, concordance rates arent 100%
EXAMPLES NOT IN PSYCH:
-Psychodynamic > cannot be falsified because explanation cannot be tested as its in the unconscious
-Humanism > cannot be falsified because self-esteem cannot be measured.
Replicability
DEFINITION:
-extent to which findings and procedures can be repeated by other researchers
EXAMPLE IN PSYCH:
-Behaviourist > replicable research eg Skinners rats to support operant conditioning > scientific method means it can be recreated.
-Cognitive > replicable research in memory for an example leading questions study was a lab exp
EXAMPLE NOT IN PSYCH:
-Psychodynamic > cannot be replicated because methods are not scientific + use case studies
Type I error
- When the null hyp is rejected + alternative accepted when it should’ve been the other way round > null hyp is actually true.
- More likely if the significance level is too lenient + high like 10% instead of 5%.
- Chance of getting type I error is the same as significance level (10% significance = 10% chance)
- Use 5% to balance between type I and II errors + reduce risk of both of them.
Type II error
- When null hyp is accepted and alternative rejected when it should have been the other way round > alternative hyp is actually true
- More likely if the significance level is too strict + low > eg more likely at 1% level than at 5%
- Use 5% to balance between type I and II errors + reduce risk of both of them.
Sections of a scientific report
- Abstract
- Introduction
- Method
- Result
- Discussions
- References
- Abstract
- Short summary including all major elements included in report.
- Includes aims, hypothesis, methods, results + conclusions
- Used to see if full report is relevant + worth reading
- Introduction
- Literature review of area > past research/studies related to current study
- Logical progression from general to specific > ending on most relevant info relating to aims + hyp which are referred to at the end of intro
- Method
- Detailed enough to allow replication
- Includes design (experimental design + IV, DV), sample (amount, demographic info, target pop, sampling method), materials (equipments, questionnaires, word list), procedure (events in study + reference to standardisation + debriefing) and ethics (how they were addressed)
- Results
- Summary of key findings
- Descriptive statistics - eg tables, graphs, measures of central tendency or dispersion.
- Inferential stats > choice of test, calculated + critical values, significance + accepted/rejected hyp
- Raw data goes into appendix (other info at end)
- Qualitative findings including analysis of themes
