HP 350L Lecture Notes - Lecture 4: Ordinal Data, Observational Error, Statistical Conclusion Validity
Document Summary
Include standardized measures, such as ces-d, which are developed and scored according to certain rules. Self-report measures are perhaps the most common data collection instruments in behavioral science research. Approach is exploratory, allow researchers to find out if participants have any responses at all. Allow participants to use their own languages and make subjects more comfortable. Can be time-consuming for participants to answer and researchers to transcribe an code. Responses may be off-the-mark from intent of question and cannot be used. Difficult to assess reliability or consistency of responses. Efficient to administer to large groups of subjects. May limit breadth and depth of subjects" responses. The most common closed-ended measuring approach (rating scale) Three commonly used rating scales formate are: numerical, forced- choice, and graphic. Numerical scale is a rating scale that corresponds with a sequence of numbers defined by the researcher. Type of numerical scale that forces subjects into making a definite statement.