HS 2801 FINAL EXAM QUESTIONS WITH
COMPLETE ANSWERS
Surveys - ANSWER - are correlational research
- causality may not be inferred
- some research makes predictions
- predictor variables and criterion variables
- types
- questionnaires, interviews, self-report diaries
Designing Good Surveys - ANSWER 1. Consider (carefully) the research question
2. Carefully define all constructs
3. Review existing instruments
4. write items for each construct
5. Seek advice on items from colleagues
6. Pilot test items
7. Analyze item statistics, check factor structure
8. Re-work items that are poor
9.. Administer final survey
Defining Constructs - ANSWER - research has shown that this the most important stage
in survey design
, - inexperienced item writers generate items that are as good as experienced item
writers if they are aware of the principles it entails and have a good definition of the
constructs
- also what demographics do you need?
type of questions - ANSWER - demographic information
- open-ended items
- no specific set of alternative answers
- critical incident technique
- analyzed with content analysis
- close - ended (restricted) items
- specific set of alternative answers
- partial open-ended items ("other" category)
- forced choice items
- rating scales
demographics - ANSWER - trend: collecting ordinal data, rather than interval data
- ex. age
- when possible, collect interval data
- easier to analyze with more robust conclusions
- only resort to ordinal data for questions about which a participant may be unsure
about specifics
, - ex. family income
open-ended questions - ANSWER - used more in interviews than questionnaires
- may be difficult to focus respondents
- critical incident technique, often require follow-up questions for clarification
- tend to be subjective
- usually analyzed using content analysis
- some open-ended questions (e.g. projective tests) are scored according to
pre-determined scoring guidelines
Rorschach Ink Blot (what do you see?) - ANSWER - the most widely used projective test,
a set of 10 inkblots, designed by Hermann Rorschach; seeks to identify people's inner
feelings by analyzing their interpretations of the blots
Thematic Apperception Test (TAT) - ANSWER - a projective test in which people express
their inner feelings and interests through the stories they make up about ambiguous
scenes
- What is happening, who are these people, has led up tot this situation, are they
thinking? why do they want, will happen, will they do?
Scales - ANSWER - categorical scales (nominal)
- continuous scales (interval/ratio)
- ranked scales (ordinal)
, - summative scales
- scaling models
categorical scales (nominal) - ANSWER participant assigns self to a category
- categories must be mutually exclusive
- e.g. gender, diagnosis, religion, etc.
continuous scales (interval/ratio) - ANSWER - e.g. blood pressure, height, age
ranked scales (ordinal) - ANSWER e.g. pain/symptom severity
cumulative scales - ANSWER each item represent an increasing amount of the attribute
being measured
summative scales - ANSWER - each item contributes equally to the total score
scaling models - ANSWER - Likert scales, the semantic differential, visual analogue scales,
Gutman scales
Likert Scales - ANSWER - ordinal-level scales containing seven points on an agree or
disagree continuum
- summative calling method involving ranked values
COMPLETE ANSWERS
Surveys - ANSWER - are correlational research
- causality may not be inferred
- some research makes predictions
- predictor variables and criterion variables
- types
- questionnaires, interviews, self-report diaries
Designing Good Surveys - ANSWER 1. Consider (carefully) the research question
2. Carefully define all constructs
3. Review existing instruments
4. write items for each construct
5. Seek advice on items from colleagues
6. Pilot test items
7. Analyze item statistics, check factor structure
8. Re-work items that are poor
9.. Administer final survey
Defining Constructs - ANSWER - research has shown that this the most important stage
in survey design
, - inexperienced item writers generate items that are as good as experienced item
writers if they are aware of the principles it entails and have a good definition of the
constructs
- also what demographics do you need?
type of questions - ANSWER - demographic information
- open-ended items
- no specific set of alternative answers
- critical incident technique
- analyzed with content analysis
- close - ended (restricted) items
- specific set of alternative answers
- partial open-ended items ("other" category)
- forced choice items
- rating scales
demographics - ANSWER - trend: collecting ordinal data, rather than interval data
- ex. age
- when possible, collect interval data
- easier to analyze with more robust conclusions
- only resort to ordinal data for questions about which a participant may be unsure
about specifics
, - ex. family income
open-ended questions - ANSWER - used more in interviews than questionnaires
- may be difficult to focus respondents
- critical incident technique, often require follow-up questions for clarification
- tend to be subjective
- usually analyzed using content analysis
- some open-ended questions (e.g. projective tests) are scored according to
pre-determined scoring guidelines
Rorschach Ink Blot (what do you see?) - ANSWER - the most widely used projective test,
a set of 10 inkblots, designed by Hermann Rorschach; seeks to identify people's inner
feelings by analyzing their interpretations of the blots
Thematic Apperception Test (TAT) - ANSWER - a projective test in which people express
their inner feelings and interests through the stories they make up about ambiguous
scenes
- What is happening, who are these people, has led up tot this situation, are they
thinking? why do they want, will happen, will they do?
Scales - ANSWER - categorical scales (nominal)
- continuous scales (interval/ratio)
- ranked scales (ordinal)
, - summative scales
- scaling models
categorical scales (nominal) - ANSWER participant assigns self to a category
- categories must be mutually exclusive
- e.g. gender, diagnosis, religion, etc.
continuous scales (interval/ratio) - ANSWER - e.g. blood pressure, height, age
ranked scales (ordinal) - ANSWER e.g. pain/symptom severity
cumulative scales - ANSWER each item represent an increasing amount of the attribute
being measured
summative scales - ANSWER - each item contributes equally to the total score
scaling models - ANSWER - Likert scales, the semantic differential, visual analogue scales,
Gutman scales
Likert Scales - ANSWER - ordinal-level scales containing seven points on an agree or
disagree continuum
- summative calling method involving ranked values
