Psychometric testing of Cooper Parental Self-Efficacy Scale-Child Health Behavior

Abstract

<p>Although numerous self-efficacy scales grounded in Bandura&rsquo;s theory of self-efficacy were available, no specific scale that measured the concept of parental self-efficacy to influence child health behavior existed.&nbsp; The researcher developed the Cooper Parental Self-Efficacy Scale-Child Health Behavior (CPSS-CHB) and previously conducted a content validation study. &nbsp;The aims of this study were to 1.) ensure that the 28-item CPSS-CHB was readable, 2.) to determine reliability of the scale, and 3.) to initiate the establishment of validity.</p> <p>Readability was established during a feasibility study with 20 participants recruited via emails and Facebook social network messages. &nbsp;Using SurveyMonkey.com, participants were directed to a link that opened an Internet-based survey.&nbsp; Participants were asked to complete the 28-item CPSS-CHB and provide basic demographic information.&nbsp; Areas for participant comments followed each item.&nbsp; The survey closed at 20 participants, although 50 contacts were made; these 20 surveys were completed within 24 hours.&nbsp; All 20 participants were parents.&nbsp; Participants included 13 (65%) women and 17 (85%) Caucasians.&nbsp; Fifteen (75%) of the participants were married, and 12 (60%) had children between 3 and 5 years of age. &nbsp;Nine (45%) participants had incomes of over $85,000 for the previous year. &nbsp;Parents ranged in age from 27 to 48 years with a mean of 36 years. &nbsp;The mean years in school was 16.&nbsp; The scale was deemed readable and easy to understand.</p> <p>In a full study, reliability was examined using 298 participants, while concurrent and convergent validity were examined using 291 participants, and 290 participants for discriminant validity.&nbsp; The participants for the reliability and validity estimation portion of the study were recruited via a similar recruitment procedure as in the readability study.&nbsp; The first 300 parents and caregivers of children ages 3 to 16 years who responded from the recruited pool of approximately 800 e-mail messages, Facebook messages, Twitter postings, and additional posts to applicable parent-focused discussion board and chat room (ExpressiveParents.com) messages were included in the study. &nbsp;The message and survey link was used to request that those in receipt of the message and survey forward the link to contacts they knew who had children between 3 and 16 years of age.&nbsp; The instrument packet contained the Cooper Parental Self-Efficacy Scale-Child Health Behavior (CPSS-CHB), General Self-Efficacy Scale (GSES) (Schwarzer &amp; Jerusalem, 1995), the Rosenberg Self-Esteem Scale (RSES) (Rosenberg, 1965) and demographic items.&nbsp; The GSES was used to examine concurrent and convergent validity.&nbsp; The RSES was used to examine discriminant validity.<strong>&nbsp; </strong>The investigator set SurveyMonkey.com for a cut off of 300 participants.&nbsp; The survey was available and administered through the World Wide Web at the Survey Monkey portal, an on-line survey web site, for seven days, at which time 300 completed surveys had been collected.&nbsp; Participants were primarily married (87.9%), Caucasian (92%) women who were parents of pre-school and school-age children. &nbsp;The median number of children in each home was two, while the mean age of the participants was 40 years (<em>SD</em> &plusmn;9.4) with mean of 16.76 years in school (<em>SD</em> &plusmn;3.1). &nbsp;Five participants acknowledged an annual income level below $15,000, while 130 participants indicated annual income levels over $85,000.</p> <p>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Using the alpha coefficient, the internal consistency reliability for the entire scale was found to be .96.&nbsp; Principal components analysis (PCA) with oblique rotation was used to determine that a three-factor structure solution was the best fit for the scale. Factors were labeled problem times, stressful times, and good times. &nbsp;Only one item did not load on a factor.&nbsp;</p> <p>With regard to concurrent validity, only a weak correlation (<em>r</em>= .17) was found between the CPSS-CHB scale and the General Self-Efficacy Scale.&nbsp; Therefore, concurrent validity was not established in this study.&nbsp; Convergent validity was not supported in view of the weak correlation of the scale with the General Self-Efficacy Scale.&nbsp; However, discriminant validity was supported by the weak correlation (<em>r</em>= .07) found between the CPSS-CHB scale and the Rosenberg Self-Esteem Scale, even though a higher correlation of .3 or .4 was expected.&nbsp;</p> <p>The CPSS-CHB was revised to include 27 items composed of three factors (problem times, stressful times, good times), judged internally consistent for this sample, but validity remains questionable.&nbsp; Construct validation of the CPSS-CHB was not demonstrated. &nbsp;Continued refinement of the scale and further validity testing is mandatory.</p>