Cognitive Testing of Human Papillomavirus (HPV) Survey Items

Purpose: Cervical cancer can be prevented by a vaccine against types of HPV that cause most of these cancers. Many studies have assessed factors associated with HPV vaccine acceptability and uptake, but few have examined whether people understand survey items that assess these potential correlates. We sought to develop and cognitively test survey items assessing constructs deemed important to vaccine acceptability.

Methods: Investigators from seven research centers across the US worked collaboratively to develop and cognitively test survey items among racially and ethnically diverse parents (n=62) of girls between ages 9-17. We used a think-aloud protocol and semi-structured verbal probing in qualitative in-person and telephone interviews. We cognitively tested items through a series of iterative phases, initially testing items with a small sample of parents, utilizing feedback to revise and re-administer to a new sample four times.

Results: The final survey contained 20 items on attitudes and beliefs relevant to HPV vaccine. Some parents misinterpreted statements about hypothetical vaccine harms as statements of fact. Some were unwilling to answer items about perceived disease likelihood and perceived vaccine effectiveness, because they said the items appeared to have a “right” answer that they did not know. Cognitive testing results led us to revise 14 questions to improve clarity and comprehension.

Conclusions: Cognitive testing of HPV vaccine survey items revealed important differences between our intended meaning and meaning ascribed to items by participants. Employing cognitive testing in an integrated strategy for instrument development can identify problems with measures and may increase measurement validity.