Date of Award:


Document Type:


Degree Name:

Doctor of Philosophy (PhD)


Special Education and Rehabilitation


Phyllis R. Publicover


The purpose of the study was to develop a valid and reliable instrument designed to assess the parenting and child rearing attitudes of adolescents, Responses to the instrument would allow for the identification of "high risk" adolescents, i.e., those adolescents whose attitudes towards parenting and child rearing indicated a high potential for child abuse.

A review of the literature identified the following four parenting constructs most commonly associated with abusive parents: Construct A: Inappropriate parental expectations of the child; Construct B: Inability of the parent to be empathically aware of the child's needs; Construct C: Strong parental belief in the value of punishment; Construct D: Role reversal. These constructs served as the basis for the development of the item domain.

The sample used in this study consisted of 2,628 adolescents attending schools located in Utah and Idaho. Employing a Likert format, three Prototypes of the instrument were developed and field tested. Substantial content validity was acquired through the judgments made by a panel of experts in child abuse, test construction, and attitudinal measurement. Construct validity of the instrument was established through the results obtained from factor analysis, interitem correlations, and item-construct correlations.

The data generated from the factor analysis indicated 32 items had the highest positive factor loadings (>= .20) in each of the four identified constructs. The range of item-construct correlations for the 32 items (.53 to .75) indicated adequate to high degrees of relationship between the item scores and total construct scores. The data obtained relative to the internal consistency of the items indicated adequate level s of reliability for each construct (Construct A, .70; Construct B, .75; Construct C, .81; Construct D, .82). The test-retest reliability coefficient of the instrument showed an adequate level of stability over a one week period of time (.76).

Raw scores, converted into factor scores, indicated approximately 3% of the adolescents who responded to the instrument scored -2 and -3 standard deviations away from the mean. The results of the multi-variant analysis of variance indicated that abused adolescents scored significantly lower (p.