Criteria For Discriminant Validity Research Articles

Developing and testing an instrument for identifying performance incentives in the Greek health care sector.

BackgroundIn the era of cost containment, managers are constantly pursuing increased organizational performance and productivity by aiming at the obvious target, i.e. the workforce. The health care sector, in which production processes are more complicated compared to other industries, is not an exception. In light of recent legislation in Greece in which efficiency improvement and achievement of specific performance targets are identified as undisputable health system goals, the purpose of this study was to develop a reliable and valid instrument for investigating the attitudes of Greek physicians, nurses and administrative personnel towards job-related aspects, and the extent to which these motivate them to improve performance and increase productivity.MethodsA methodological exploratory design was employed in three phases: a) content development and assessment, which resulted in a 28-item instrument, b) pilot testing (N = 74) and c) field testing (N = 353). Internal consistency reliability was tested via Cronbach's alpha coefficient and factor analysis was used to identify the underlying constructs. Tests of scaling assumptions, according to the Multitrait-Multimethod Matrix, were used to confirm the hypothesized component structure.ResultsFour components, referring to intrinsic individual needs and external job-related aspects, were revealed and explain 59.61% of the variability. They were subsequently labeled: job attributes, remuneration, co-workers and achievement. Nine items not meeting item-scale criteria were removed, resulting in a 19-item instrument. Scale reliability ranged from 0.782 to 0.901 and internal item consistency and discriminant validity criteria were satisfied.ConclusionOverall, the instrument appears to be a promising tool for hospital administrations in their attempt to identify job-related factors, which motivate their employees. The psychometric properties were good and warrant administration to a larger sample of employees in the Greek healthcare system.

Consumption Communication and Parental Control of Children’s Television Viewing: A Multi-Rater Approach

The increasing influence and involvement of children and teenagers on various aspects of consumer behavior is well established. Past studies have examined consumer socialization primarily from the perspective of the mother (e.g., Carlson, Grossbart, and Walsh 1990; Rose, Bush, and Kahle 1998; see Palan 1998 for an exception). This study extends this research by examining the relationship between family communication and parental control over children’s television viewing among multiple family members (the mother, the father, and the child). The results generally support the convergent validity of the measures across raters. The criteria for discriminant validity (Campbell and Fiske 1959) were generally met, but some complexities were found, particularly for the ratings or perceptions of the child.

Reliability, validity, and responsiveness of general and disease-specific quality of life measures in a clinical trial for cytomegalovirus retinitis

The objective of this study was to evaluate a questionnaire for assessing general and disease-specific quality of life among people with cytomegalovirus (CMV) retinitis. Cross-sectional and longitudinal analyses of data from 279 people enrolled in the CMV Retinitis Retreatment Trial were used. At baseline, Cronbach's alpha and multitrait analysis were used to assess internal consistency and discriminant construct validity for scales of general health, vision, and treatment impact. Associations of scales with clinical measures of health and vision were assessed at baseline with Pearson correlations and t tests, and over time with generalized estimating equations regression. Internal consistency coefficients ranged from .68 to .88. Criteria for discriminant validity were fulfilled for most scales; however, the general health perceptions and energy scales were highly correlated. Scales were moderately correlated with clinical measures at baseline. Over time, scale scores were associated with Karnofsky scores and clinical measures of CMV retinitis and vision. General and CMV retinitis-specific quality of life measures appear reliable, valid, and responsive.

The Construct Validity of Task Inventory Ratings: A Multitrait-Multimethod Analysis

This study investigated the construct validity of task inventory ratings from job incumbents and supervisors in 2 jobs on the 4 task dimensions of importance, criticality of error, difficulty to learn, and time spent. Job incumbents completed task ratings on either the typical relative scales or absolute scales developed for each dimension; supervisors completed relative ratings only. Relative ratings from incumbents and supervisors were each compared in multitrait-multimethod analyses with the absolute ratings. Ratings on all task dimensions demonstrated high convergent validity, and time spent and difficulty to learn ratings met criteria for discriminant validity in at least 1 job. Practical implications regarding redundancy among ratings on different task dimensions, the use of both incumbents and supervisors as job analysis respondents, and the use of absolute scales in task analysis are discussed.

Construction of scales for the Center for Marital and Sexual Health (CMASH) sexual functioning questionnaire

The objective of this study was to identify a set of scales for summarizing the results of the Center for Marital and Sexual Health Sexual Functioning Questionnaire (CMASH SFQ). Scales for this instrument were constructed using patients' responses to the CMASH SFQ in a recent clinical trial of prostaglandin E1 (PGE, alprostadil), an injectable vasodilator used to treat erectile dysfunction. A set of items was identified as a scale if they met predetermined standards of internal consistency, discriminant validity, and convergent validity. The reproducibility of the scales was then evaluated, and the extent to which the scales reliably measured change in treatment-related quality of life was assessed. Reliable and responsive scales for four domains related to sexual functioning were identified. Items in these scales were generally internally consistent; items in the scales and the scores for the scales generally satisfied criteria of discriminant validity; and the scales generally satisfied the standard for convergent validity. Evaluating patients' responses to the CMASH SFQ with scores for four reliable and responsive scales allows easier summarization of patients' sexual functioning and increases the statistical power of evaluations of the effects of interventions designed to improve sexual functioning .

Convergent and Discriminant Validity of the Borderline Syndrome Index

The convergent and discriminant validity of the Borderline Syndrome Index was examined by using a multitrait-multimethod matrix 80 subjects with DSM-III—R diagnoses of Borderline Personality Disorder completed three self-report measures, namely, the Borderline Syndrome Index, the MMPI Schizophrenia subscale, and the Beck Depression Inventory. A semistructured interview was blindly conducted with each subject, and symptom ratings were made on the Diagnostic Interview for Borderlines, New Haven Schizophrenia Index, and Hamilton Rating Scale for Depression. High convergent validity was found on the Borderline Index and the Diagnostic Interview for Borderlines. While fulfilling most of the criteria for discriminant validity, the Borderline Syndrome Index scores were also significantly correlated with depression scores on the Beck and Hamilton measures of depression.

Measuring Components of Childrenʼs Health Status

Measures of physical, mental and social components of health status and general health ratings were studied for children ages 0-4 (N = 679) and 5-13 (N = 1473). Questionnaires were completed by adult proxies (usually mothers) in three generally healthy populations. Hypothesized multi-item scales were tested; reliability was estimated and preliminary attempts at validation were undertaken. Items in ten scales pertaining to mental health (Anxiety, Depression, Positive Well-Being, Mental Health Index), social health (Social Relations), general health ratings (Current Health, Prior Health, Resistance/Susceptibility to Illness, General Health Rating Index), as well as parental satisfaction with child development satisfied Likert-type and discriminant validity criteria. Because functional limitation items were endorsed for very few children, scales to measure physical health could not be tested. Almost all scales were sufficiently reliable for group comparisons; reliability coefficients were lower in the most disadvantaged population. Interrelationships among scales and validity variables generally supported their construct validity and supported a multi-component model of children's health status.

Convergent and Discriminant Validities of Two Sets of Measures of Spatial Orientation and Visualization

Responses from high school sophomores to the Eliot-Price and the Guilford-Zimmerman tests of spatial orientation and visualization were evaluated in terms of the multitrait-multimethod matrix technique. The Eliot-Price tests were found to meet all criteria for convergent and discriminant validity. However, the Guilford-Zimmerman tests did not meet one of the criteria for discriminant validity. The Eliot-Price tests appeared to be more nearly precise measures of spatial orientation and visualization.