Research Ch. 13 - Scrutinizing Data Collection Methods
Terms
undefined, object
copy deck
- 3 major types of data used in QUANTITATIVE Research:
-
1. Self-Report
2. Observation
3. Biophysical
*OPERATIONALIZATION of the variables is decided during the PLANNING phase - Self-Report Methods:
-
- most common form of data collection
- structured interviews or questionnaires
- can gather RETROSPECTIVE data
- serious concern is VALIDITY and ACCURACY of reports - Types of Questions
-
1. Open ended: response in own words
- more difficult to administer and analyze
- respondent might be reluctant to provide written responses; or unable to do so
- allows for rich and full responses
2. Closed ended: response alternatives are pre-specified by the investigator
- allows for comparability of responses among participants
- difficult to construct
- easy to administer
- easy to analyze
- efficient
- might over-look important potential responses
- might be superficial
- respondents could be forced to choose a response that does not fit their view/experience - Instrument Construction:
-
- careful wording of items for clarity
- counter balance positively and negatively worded item
- sensitivity to participants' views/feelings; nonjudgmental
- avoidance of bias
- attention to reading level
- critical review of items by experts/potential participants
- pretesting - Questionnaires (vs. Interviews)
-
- LESS COSTLY (people don't have to spend time for interviews; can send questionnaires by email)
- LESS TIME and MORE EFFICIENT to administer
- Can allow for ANONYMITY (unknown identities) - Interviews (vs. Questionnaires)
-
- response rate tends to be HIGHER (what is the importance of that?)
- feasible for most people; some people might not be able to complete a questionnaire
- allows for observation of participant and setting which can be useful in interpreting responses - 3 Main forms of STRUCTURED SELF-REPORT:
-
1. Likert Scale
2. Semantic Differential Scale
3. Visual Analog Scale - LIKERT SCALE
-
- summated rating scale
- used to measure PSYCHO-SOCIAL VARIABLES (ex. attitudes, perceptions, needs, etc.)
- DECLARATIVE statements about a particular topic
- Respondents usually AGREE or DISAGREE on a scale (usually 5 or 7 point)
- Scores are summed for COMPARISON among participants
- bigger range; covers more than attitudes; similar to semantic differential scale) - SEMANTIC DIFFERENTIAL SCALE
-
- used to measure ATTITUDES
- composed of BIPOLAR ATTITUDES (ex. good/bad, important/unimportant)
- 7 point scale (from one extreme of the characteristic to the other)
- respondents rate the attitude from 1-7
- scores are summed - VISUAL ANALOG SCALE
-
- used to measure SUBJECTIVE EXPERIENCES (ex. pain, dyspnea)
- 100 mm VERTICAL or HORIZONTAL LINE with DESCRIPTIVE labels at each end
- participants mark the line at the point that BEST represents the level of the experience (ex. intensity of a symptom) - DISADVANTAGES of SCALES:
-
- SOCIAL DESIRABILITY response set bias - tendency to provide socially desirable responses
- EXTREME response set bias - tendency to provide extreme responses
- ACQUIESCENCE response set bias - tendency to always AGREE or DISAGREE with statements - How to Critique Self-Reports:
-
- difficult to critique self-reports in publications because often detailed description is not provided
- did the researcher use the best mode of data collection given respondent characteristics and the research questions (ex. structured interview, self-administerd questionnaire, etc.)
- Does the instrument take into account the complexity of the problem under investigation? - Observation (QUANTITATIVE)
-
- Structured
- CATEGORIES and CHECKLISTS for documenting the behaviours
- could also use a RATING SCALE
- TIME SAMPLING: selection of specific time periods for sampling observations
- EVENT SAMPLING: select behaviours for observation
- suited to situations where people cannot/or prefer not to describe their behaviours (ex. children, persons with mental illness, sensitive issues)
- possible ethical difficulties & reactivity
- OBSERVER bias
- quality of the data is DEPENDENT on the skills of the observer (observational and interpersonal) - How to Critique QUANTITATIVE Observation?
-
- was observation an appropriate approach given the research question?
- was the observation procedure well described (ex. Who observed? What was observed? Was there concealment? Event or Time sampling? What steps were taken to reduce observer bias?
- How were the data recorded? - BIOPHYSICAL MEASURES:
-
- can be IN VIVO (ex. BP, PFT, MRI) or IN VITRO (ex. blood test, pathology test)
- ACCURATE and PRECISE
- OBJECTIVE
*in vivo (taking place IN an individual)
*in vitro (taking place OUTSIDE an individual - ex. test tube, culture dish, etc.) - How to Critique BIOPHYSICAL MEASURES?
-
- Is the measure appropriate for the RESEARCH QUESTION?
- Was CARE taken during the measurement to ensure ACCURACY?
- Were the findings INTERPRETED ACCURATELY? - IMPLEMENTING the data collection plan:
-
- WHO will conduct the data and are they QUALIFIED and properly trained?
- WHERE were the data collected and under what CIRCUMSTANCES?
- HOW were the data collected?