RM

43 flashcards · Other · by Daniel F.

1. steps of research process

   problem statement research question - fact oriented information gathering question literature research, expert insights - conduct inreviews, other studies to determine IV and DV conceptual model - conceptual (theoretical) and operational (practucal) def and relationship of constructs hypothesis - expectation of how IV(s) influence DV indicators - how do I measure the construct research instrument, questions - make a question out of the indicators

2. Issues with research to account for

   validity reliability respondent error

3. Measurement errors

   Measurer - e.g. the way you ask the questions participant - e.g. scared or embarrassed measurement instrument - e.g. only positive options for answers

4. Validity and types

   Do I measure what I want to measure internal validity external validity

5. Internal validity and forms

   the ability of a research instrument to measure what you want to measure content validity face validity criterion related validity predictive validity concurrent validity construct validity convergent validity discriminant validity

6. External validity

   extent to which findings can meaningfully be generalized to other settings, situations and/or persons

7. Content validity

   extent to which the measuring instrument provides adequate coverage of the concept you study how to get: Use literature or outside expert to raise concept validity

8. Face validity

   ‘at first sight’ a good translation of the construct how to get: select groups of experts to give their opinion

9. Predictive validity (criterion)

   ability of test to predict a future outcome based on a well established measure (criterion), measured in different points of time

10. concurrent validity (criterion)

    ability of a test to predict the present, based on a well established measure (criterion), measured at the same time

11. convergent validity (construct)

    measures of constructs that theoretically should be related to each other are, in fact, observed to be related to each other

12. discriminant validity (construct)

    measures of constructs that theoretically should not be related to each other are, in fact, observed to not be related to each other

13. Reliability and forms

    the extent to which a measure provides consistent results More reliable = less random measurement error Stability (fluctuations in results because of personal and situational aspects changing over time) Equivalence (fluctuations in results because of difference between researchers at the same point in time) → standardized forms fix this Internal consistency (degree to which multiple items measure the same construct) Split-half method: choose one half of the items randomly and compare those to t…

14. Nominal data type

    Number indicate categories, order is not meaningful

15. Dichotomous data

    Nominal data - yes or no

16. Ordinal data type

    Number indicate categories order IS meaningful e.g. clothing size

17. Interval data type

    Order in the number is important intervals between numbers are fixed ARBITRARY ZERO (only + and - ) e.g. temperature scale or iq scale (can do calculations)

18. ratio data type

    Order in the number is important fixed intervals between number meaningful zero point

19. Internal data type questions forms

    for it to be internal it needs to be symmetric perfect antonyms odd number of answers same distance between answers

20. Mistakes when making questions

    should it be asked is it comparable is it meaningful can the participant understand can the participant answer does the participant want to answer is it essential

21. how to ask for sensitive information on a measuring instrument

    projective questions - what do other people think about this topic? Or welcome participants by acknowledging both sides of the coin (some people say this some people say that)

22. Mistakes when making questions

    not enough context given participant doesn't understand and guesses too direct and sensitive has to be projective question double barreled question one question per question leading question rephrase to no lead e.g. just like other people ambiguous question make it more informal so participant understands make it precise double negative too positive or too negatively phrased question implicit positivity bad question ordering funnel approach should be used, from broad to specific

23. consequence of bad measurement instrument

    Impossible to use the data Invalid and useless conclusions and implications

24. Open questions

    good for exploration

25. Closed questions

    limit how participants can answer dichotomous scales categorical scales

26. Good answers for closed questions

    need to be mutually exclusive and collectively exhaustive bipolar adjectives are contradictions higher number indicate better balanced answering range, negative and positive specific: 4-3 times instead of "often" sensitive topics should not have too direct answers Don't know options by facts and not by optinion ranking scales should be paired, rank order, constant sum

27. create the MC question based on

    pilot research focus groups previous open questions

28. categories of people when sampling

    population - all members sampling frame - everyone who can be reached sample - selection of members

29. systematic variance of samples

    whatever sample you draw, it can never be representative

30. Sampling/random error

    sample does not represent population

31. Characteristics of correct sample

    accurate: no systematic bias no systematic variance precise: values in the sample similar to population no sampling error

32. Probability sampling and types

    Based on random selection: each unit has the same chance of being selected simple random sampling systematic sampling stratified sampling cluster sampling

33. simple random sampling

    Each unit has the same probability of being selected (higher than 0) Pros simple could be representative sample Cons Requires a lot information about your sampling frame (complete list) May result in samples that do not fit your research questions Potential for large errors Effortful and expensice

34. systematic sampling

    Selects every n-th element of the population from a random starting point Pros simple Cons Requires a lot of information about sampling frame If population list contains trend, result may be biased

35. stratified sampling

    Divide populations into sub-populations, self selected(=strata) Statistical advantage: if strata are large enough, you can test differences between strata Ideally: large differences between strata, small differences within strata

36. cluster sampling

    Population consist of clusters of elements which are close to each other, already existing E.g. random choice of cities and take a sample from supermarkets Only supermarkets from the selected clusters are studied

37. non probability sampling and types

    Not based on random selection: different selection probabilities for the elements More subjective than probability sampling Selection probability is unknown Convenience sampling judgment sampling quota sampling snowball sampling

38. Convenience sampling

    Freedom to choose whoever can be found or wants to participate Quick and dirty Can be useful to test an idea in limited time

39. Snowball sampling

    Respondents identify other potential respondents for the researcher Efficient way to find respondents in niche situations (croquet players) Good chance that respondents know similar respondents sample not representative for the population

40. Sample size dependent on (needs to bigger if)

    high variation of population emphasis on subgroups precision needed high chance of non-responders

41. Non-response types

    Not present at moment of contact refusal to participate full, not interested partial, e.g. privacy concerns unable to find respondent

42. Random vs systematic non-response

    random Respondent and non respondents do not differ systematically on important variables Result can give a correct view of reality systematic Respondent and non respondents differ systematically on important variables Result can give a biased view of reality Needed: larger and better sample

43. Qualitative vs Quantitative research

    Qualitative - soft data quantitative - hard data