Karteikarten
two
Karten in diesem Stapel (29)
- 1how should you enter data from a questionnaire into SPSSenter data exactly as it appears on the questionnaire
- 2what four things should you consider when establishing a scale e.g. on a questionnaire1. deal w one (sub)scale at a time
2. decide what scale is about
3. consider name of scale
4. consider the most logical interpretation - 3when doing initial re-coding, what should all item scores beall should point in the same direction (e.g. that load positively relative to the concept)
- 4when recording recoded items, what should you do?turn negatived items to score correctly towards concept (items that load negatively relative to concept need to be recoded into a new variable)
- 5what is reliability in terms of a questionnaire?is it a meaningful tool? the quality of the instrument
- 6what is validity in terms of a questionnaire?does it test what it claims to test?
- 7two ways to measure reliability of a questionnaire?test re-test
split half - 8what is item analysissorting out the good questions from the bad questions so all questions contribute to the overall score
- 9what are two ideas about cronbach's alpha1. if the items on a questionnaire fit together coherently, alpha is close to 1.0
2. if an item makes a useful contribution to a questionnaire, the scores on it will correlate with the questionnaire total - 10if someone does not complete all of the items in the questionnaire, what do you do with their data?exclude it from the analysis
- 11when analysing a questionnaire, if aiming for alpha > .7 (e.g.) what do you do?use item total statistics. Recode one item at a time and reanalyse
- 12what is the name of the alpha to analyse these itemscronbachs alpha
- 133 potential problems with items on questionnaire?1. question may have been coded back to front,, 2. everyone might give same answer,, 3. question might have nothing to do with measure
- 14explain what it means for the question to be coded back to frontthis means the question appears to measure the opposite of what was intended. Often negative statements
- 15how can we tell if the question has been coded back to frontif the item total correlation is negative
- 16explain what it means for everyone to give the same answermeans the question does not differentiate between people and can happen if its a question that is extreme or unlikely to yield any other view
- 17how can we tell if everyone gives the same answerif the item total correlation is low or if the alpha (if deleted) is high than the current alpha
- 18explain what it means if the question has nothing to do with the measurethe question does not measure the right thing e.g. a question that is not relevant to the topic
- 19how can we tell if the question has nothing to do with the measureif the item total correlation is low or if the alpha (if deleted) is high than the current alpha
- 206 different types of validity?known groups validity, concurrent, construct, content, criterion, face validity
- 21define known groups validitydifferent scores for people known to differ?
- 22define concurrent validitynew scale cf. 'gold standard'?
- 23define construct validityconsistent with theories?
- 24define content validityall aspects for concept?
- 25define criterion validityresults consistent with other measures?
- 26define face validitydo experts/clients/participants agree that construct is being measured?
- 27what is the relationship between validity and reliabilitywithout reliability, there is no validity
- 282 examples of criterion validity?do scores for a questionnaire to measure aggressiveness correlate w convictions for violent crime?
IQ tests being designed to correlate with age in children and academic achievement - 29what is the relationship between validity and reliabilitywithout reliability there is no validity
Informationen
- Beschreibung
- PSYCHOLOGY YEAR 2 (2 rmd) Karteikarten am two, erstellt von Ki Rai am 22/01/2019.
- Erstellt von
- Ki Rai
- Aufrufe
- 2
- Bewertung