- Posts: 4
Great question. In the ideal setting, children will read and answer the questions on their own in a private location.
However, some kids cannot read the questions but could answer the questions if the questions were read to them. in this circumstance, if a computer-generated voice could read the questions and allow the child to say their response (or use a touch screen), that would be best. In other circumstances, someone may need to read the question. Any time, a person is involved in the process, there is potential for bias. In a study by Magnus et al 2016, we found negligible differences between self-report (via computer) and interviewer-administered PROMIS Pediatric questionnaires. These interviewers had no relationship with the child.
When a parent/caregiver becomes involved in the assessment, there is greater potential for bias. The parent may unconsciously insert their beliefs about the child's responses in the assessment. (e.g., isn't it true, Alex (made up name), that sometimes you feel sad?). Of course, when this happens, then there are concerns about the validity of the data. Thus, one would have to interpret those data with caution. For research, it would be good to note when this occurs. There are examples of surveys that include a question at the end that asks if anyone helped the patient complete the questionnaire. For clinical care, it may be worth for the clinician to confirm some of the answers with the child.
Hope this is helpful.
There is a naturally occurring phenomenon in our clinic where the parents want to either read the questions to their kids, or help them answer the pediatric (not proxy!) surveys. How does this impact my scores and interpretation, both scientifically and clinically?