[Construction and initial validation of a research attitude questionnaire in primary care]

To construct and validate a research attitude questionnaire in primary care physicians.

Validation study of a attitude questionnaire.

Teaching health centres of family medicine in Valencia (Spain).

All the physicians (253) of 18 centres: family and general physicians, pediatricians and third-year family residents.

It was constructed a five Likert scale of 31 items and two dichotomic items. Pretest phase to 25 physicians. Validation Phase: 1. Reliability: a) Spearman-Brown correlation coefficient item-global and Cronbach coefficient; b) test-retest through comparing item punctuation and global. 2. Content validation; construction validation through exploratory factorial analysis and method of extreme groups (research or no) [corrected]. One hundred and fifty physicians (59.3%) answered the first contact, and 107 (71.3%) of them, the second contact. The questionnaire evidenced good internal consistency (alpha = 0.82) and time stability. Five factors were extracted that explained 58.7% of variability: research integration in primary care, self-capacity in study development, subjectives barriers, perception of education's need in research methodology, and structural barriers. Three factors showed significant differencies between extreme groups [corrected]. Definitive questionnaire of 20 items.

The questionnaire is a valid and reliable instrument to know the research attitude of primary care physicians.