Table 2

PAHD Scale development
Item stem identification A four-member research team was responsible for item generation and refinement. Initially, the team became immersed in the data matrices of the struggling to adjust construct. Independent raters created a profile of frequency and priority ratings of construct properties and descriptors (e.g., dwelling on carrier status, positive outlook, concern for young family members, importance of openness, strain on relations, emotional burden of suffering & death) by participant and group. Team members used these profiles to generate item stems for 5 groups and the principal investigator validated the process. At this stage, the team had 59 potential items.
Item stem reduction Multiple drafts of items for the scale were reviewed and modified by the researchers. Team meetings were held frequently to collate, prioritize and refine item stems for potential scale inclusion (emphasis on conciseness, avoidance of negative wording, ambiguous terminology, jargon, value-laden words and double-barreled questions). A final set of 17 items were identified for potential inclusion in the PAHD scale.
Rating scale development Initial rating scales focused on the frequency of occurrence (never, rarely, sometimes, often, or almost always), and ‘the importance/difficulty/receptiveness of’ or ‘how satisfied/concerned/confident/certain one was with’ select events/situations (not at all, a little bit, moderately, quite a bit, extremely). The multiple selection options made things cumbersome and confusing. The decision was made to rework the items and use one rating scale. Despite recognizing that a 5-point scale might not be sufficient for maximum reliability, the group consensus was that it would be difficult to devise unambiguous additional ordinal adjectives.
Scale readability Several tools (i.e., Flesch-Kincaid Grade Level and Flesch Reading Ease, Fog index and SMOG) were used to assess the PAHD’s reading level at less than or equal to Grade 10. Although a grade less than 10 is recommended to ensure maximum reading ease and material comprehension, the PAHD is developed to assess the experiences of individuals who have had predictive DNA testing. These individuals have had repeated exposure to terms such as LS, hereditary non-polyposis colorectal cancer, carriers/non-carriers, inherited, generations, genetic and geneticist/genetic counselor. These polysyllabic words and others are used frequently throughout the scale which does increase the final readability score.
Content validation First, two genetic counselors (GCs) who work with individuals during the genetic testing process reviewed the PAHD. A brief written synopsis of the conceptual model and construct definitions, along with a copy of the scales, were given to the GCs to prepare them for this task. Input was requested on item content relevancy (extremely, moderately, slightly, or irrelevant) in terms of its ability to measure the properties of targeted constructs, and effectiveness (very, moderately, poorly or not at all effective) of the 5-point Likert rating scale for ease of item rating. Minor changes to select items were made based on their recommendations. Second, the PAHD was administered to individuals (carrier & non-carrier) who had participated in the survey and qualitative studies. Respondents were asked to comment on item clarity/relevancy, and rating scale usefulness. No changes were made at this stage.

Watkins et al.

Watkins et al. BMC Psychology 2013 1:7   doi:10.1186/2050-7283-1-7

Open Data