HLTH 8202E Week 7 Response Scale Review Example

Response Scale Review: Every Option Set in the Home Health Aide Heat Survey Checked for the Respondent It Leaves Without a True Answer, With a No-Visits Category, Labeled Frequency Points and Agreement Scales Replaced

Student Name

Walden University

HLTH 8202E: Survey Research Methods

Instructor Name

Month Day, Year

Response Scale Review: Every Option Set in the Home Health Aide Heat Survey Checked for the Respondent It Leaves Without a True Answer, With a No-Visits Category, Labeled Frequency Points and Agreement Scales Replaced

The Test

A question can be well worded and still produce a wrong answer if its options leave no honest place for some respondents. This review goes through each scale in the draft survey and asks one question: what would an aide in an unusual situation be forced to choose? Each scale is shown as drafted and as revised, with the reason for each change.

Exposure Frequency

Drafted: never; sometimes; often.

Revised: I did not visit any clients that summer; no visits felt that hot; a few visits; around half; the majority; every visit or close to it.

Reason: The drafted scale had three problems. An aide who made no visits last summer, because of leave or a new job, would pick "never" and look unexposed, which would lower the exposure estimate. The new first option removes that case from the denominator. Second, "sometimes" and "often" mean different things to different respondents; anchoring each point to a share of visits gives every aide the same reference. Third, the scale lacked the extremes, so aides who were hot on every visit had no way to say so.

Control Over Cooling

Drafted: yes; no.

Revised: usually yes; I had to ask first; the client or family made that choice; it differed too much between homes to say.

Reason: Aides work in someone else's home. Many could ask to open a window but would not, because of the client's wishes or the relationship. A yes-or-no item would force those aides to choose between two answers that are both partly false. The added options separate permission from availability and give a place to aides whose situation varied by client.

Symptom Frequency

Drafted: rarely; sometimes; frequently.

Revised: not at all; one or two times over the summer; several times a month; roughly weekly; more often than weekly.

Reason: Vague quantifiers are interpreted relative to what the respondent thinks is normal, so an aide who works in heat every day may call weekly headaches "rarely." Labeling every point with a frequency reduces that drift. Five points cover the plausible range without asking aides to distinguish between frequencies they cannot recall precisely.

Training Items

Drafted: four statements, each with strongly agree to strongly disagree, such as "My training prepared me to recognize heat illness."

Revised: item-specific questions, such as "Did your training cover how to recognize signs of heat illness?" with yes, no and I don't remember, followed by "How useful was that part of the training during last summer's visits?" with very useful, somewhat useful, not very useful and not at all useful.

Reason: Agreement scales pull some respondents toward agreeing whatever the statement says, especially among respondents who are tired, less engaged or less comfortable with the survey's language (Krosnick, 1991). Asking directly about what the training covered, and then about its usefulness, gives answers that mean the same thing for every respondent.

Midpoints and "Don't Know" Options

The review keeps "I don't know" for factual items where aides may genuinely not know, such as whether a home had working air conditioning in another room. It does not add a neutral midpoint to the usefulness scale. Schuman and Presser (1981) found that offering a middle alternative or an explicit no-opinion option changes the share of respondents who choose substantive answers. For a question about training the aide received, a midpoint would mainly attract respondents avoiding a judgment, and a filter question, whether the training covered the topic, already removes those who have no basis for one.

Order of Options

Frequency scales run from lowest to highest, with the no-visits option first so that aides who did not work can leave the item quickly. In the response model of Tourangeau and colleagues, the final stage is mapping a judgment onto the options offered (Tourangeau et al., 2000), and a consistent direction across items reduces mapping errors.

What the Review Changes Overall

The revised scales add categories for aides whose situations the draft ignored, replace vague words with labeled frequencies and remove agreement formats. The pretest will check whether aides use the new categories as intended, especially the no-visits option and "it varied too much to say."

References

Krosnick, J. A. (1991). Response strategies for coping with the cognitive demands of attitude measures in surveys. Applied Cognitive Psychology, 5(3), 213-236. https://doi.org/10.1002/acp.2350050305

Schuman, H., & Presser, S. (1981). Questions and answers in attitude surveys: Experiments on question form, wording, and context. Academic Press.

Tourangeau, R., Rips, L. J., & Rasinski, K. (2000). The psychology of survey response. Cambridge University Press.