Content area
Abstract
This study's purpose was to develop an instrument to measure felt stigma experienced by cancer survivors and to test its psychometric properties. The investigator developed instrument items from a qualitative study on prejudice (Fletcher, 1989), whose data described the construct of stigma rather than prejudice. From these data and a literature review, a 37 item Likert scaled instrument evolved asking subjects to reply to each statement with varying degrees of agreement. Three cancer survivors examined the instrument and stated it appeared to measure felt stigma thus establishing content and face validity. A convenience sample (n = 189), obtained from cancer clinics and support groups in a large metropolitan area, tested the psychometric properties of the Cancer Survivor Felt Stigma Scale. The criteria for subject inclusion were: (1) experienced cancer; (2) finished initial phase of intense treatment; (3) older than 18 years; and (4) actively employed or engaged in community activities. Principal component factor analysis utilizing an unrestricted analysis revealed six factors, a compression of the original ten factors suggested by Fletcher, 1989, accounted for 66% of the variance. These six factors did appear logical with the conceptual framework. The six factors were Change in Relationships, Depersonalization, Vocational Disadvantage, Insurance Difficulties, Overprotection, and Appearance Change. Analysis of the factor loadings indicated 23 items that uniquely loaded on one factor at the $>$.40 level. The Cronbach alpha of the 37 item scale was.9627 and.9287 on the 23 item revised scale. The Cronbach alpha on the subscales ranged from.7634 to.9490. Item analysis revealed means varied from 1.70 to 3.31. Using t tests, there were significant differences between this instrument's total score means of groups by site (p =.008); whether they had surgery (p =.000); and if they saw their appearance changed by cancer (p =.000). Utilizing ANOVA, the total score group means by age were significant (p =.02). Results indicate this instrument demonstrated beginning construct validity and internal consistency reliability. The establishment of reliability and validity is an ongoing process and additional psychometric testing of this instrument using larger representative samples would be essential.
