Tests of measurement invariance failed to support the application of the "then-test"

Objective
The use of then-test (retrospective pre-test) scores has frequently been proposed as a solution to potential confounding of change scores because of response shift, as it is assumed that then-test and post-test responses are provided from the same perspective. However, this assumption has not been formally tested using robust quantitative methods. The aim of this study was to compare the psychometric performance of then-test/post-test with traditional pre-test/post-test data and assessing whether the resulting data structures support the application of the then-test for evaluations of chronic disease self-management interventions.

Study Design and Setting
Pre-test, post-test, and then-test data were collected from 314 participants of self-management courses using the Health Education Impact Questionnaire (heiQ). The derived change scores (pre-test/post-test; then-test/post-test) were examined for their psychometric performance using tests of measurement invariance.

Results
Few questionnaire items were noninvariant across pre-test/post-test, with four items identified and requiring removal to enable an unbiased comparison of factor means. In contrast, 12 items were identified and required removal in then-test/post-test data to avoid biased change score estimates.

Conclusion
Traditional pre-test/post-test data appear to be robust with little indication of response shift. In contrast, the weaker psychometric performance of then-test/post-test data suggests psychometric flaws that may be the result of implicit theory of change, social desirability, and recall bias.

The reliability of two visual motor integration tests used with health adults

Occupational therapists often assess the visual motor integration (VMI) skills of children, adults, and the elderly, which are parts of the Body Functions and Structures of the International Classification of Functioning, Disability and Health. Objective. As it is imperative that therapists use tests and measures with strong psychometric properties, this study aims to examine the reliability of two VMI tests used with adults. Method. Sixty-one healthy adults, 18 males and 43 females, with an average age of 31.82 years, completed the Developmental Test of Visual Motor Integration (DTVMI) and the Full Range Test of Visual Motor Integration (FRTVMI). The Cronbach's alpha coefficient was used to examine the tests’ internal consistency, while the Spearman's rho correlation was used to evaluate the test–retest reliability, intrarater reliability, and interrater reliability of the two VMI tests. Results. The Cronbach's alpha coefficient for the DTVMI and FRTVMI was 0.66 and 0.80, respectively. The test–retest reliability coefficient was 0.77 (p < .001) for the DTVMI and 0.61 (p < .001) for the FRTVMI. The interrater reliability correlation was significant for both DTVMI at 0.79 (p < .001) and FRTVMI at 0.80 (p < .001). The DTVMI intrarater reliability correlation result was 0.95 (p < .001) and the FRTVMI at 0.87 (p < .001). Conclusion. Overall, the DTVMI and the FRTVMI exhibited moderate to high levels of reliability when used with a sample of healthy adults. Both VMI tests appear to exhibit reasonable levels of reliability and are recommended for use with adults and the elderly.

The reliability of two visual motor integration tests used with children

Occupational therapists often assess the visual motor integration (VMI) skills of children and young people. It is important that therapists use tools with strong psychometric properties. This study aims to examine the reliability of 2 VMI tests. Ninety-two children between the ages of 5 and 17 years (response rate of 31%) completed 2 VMI tests: the Developmental Test of Visual Motor Integration (DTVMI) and the Full Range Test of Visual Motor Integration (FRTVMI). Cronbach's alpha coefficient was used to examine the internal consistency of the 2 VMI tests whereas Spearman's rho correlation was used to evaluate the test–retest reliability, intrarater reliability, and interrater reliability of the 2 VMI tests. The Cronbach's alpha coefficient for the DTVMI was .82 and .72 for the FRTVMI. The test–retest reliability coefficient was .73 (p = .000) for the DTVMI and .49 (p = .05) for the FRTVMI. The interrater correlation was significant for both the DTVMI at .94 (p = .000) and FRTVMI at .68 (p = .001). The DTVMI intrarater reliability correlation result was .90 (p = .000) and the FRTVMI at .85 (p = .000). Overall, the DTVMI exhibited a higher level of reliability than the FRTVMI. Both VMI tests appear to exhibit reasonable levels of reliability and are recommended for use with children and young people.

Generic and diabetes-specific well-being in the AT.LANTUS follow-on study : further psychometric validation of the W-BQ28 indicates its utility in research and clinical practice in type 2 diabetes in the UK

Aims  To undertake further psychometric validation of the W-BQ28 to determine its suitability for use in adults with Type 2 diabetes in the UK using data from the AT.LANTUS follow-on study.

Methods  A total of 353 people with Type 2 diabetes participated in the AT.LANTUS Follow-on study, completing measures of well-being (W-BQ28), treatment satisfaction (DTSQ) and self-care (SCI-R). Confirmatory factor analyses was used to confirm the W-BQ28 structure and internal consistency reliability was assessed. Additional statistical tests were conducted to explore convergent, divergent and known-groups validity. Minimal important differences were calculated using distribution and anchor-based techniques.

Results  Structure of the W-BQ28 (seven four-item subscales plus 16-item generic and 12-item diabetes-specific scales) was confirmed (comparative fit index = 0.917, root mean square error of approximation (RMSEA) = 0.057). Internal consistency reliability was satisfactory (four-item subscales: alpha = 0.73–0.90; 12/16-item scales: α = 0.84–0.90). Convergent validity was supported by expected moderate to high correlations (rs = 0.35–0.67) between all W-BQ28 subscales (except Energy); divergent validity was supported by expected low to moderate correlations with treatment satisfaction (rs = −0.03–0.52) and self-care (rs = 0.02–0.22). Known-groups validity was supported with statistically significant differences by sex, age and HbA1c for expected subscales. Minimal important differences were established (range 0.14–2.90).

Conclusions  The W-BQ28 is a valid and reliable measure of generic and diabetes-specific well-being in Type 2 diabetes in the UK. Confirmation of the utility of W-BQ28 (including establishment of minimal important differences) means that its use is indicated in research and clinical practice.

Reliability of performance in repeated sprint cycling tests

We have estimated the reliability of performance in a commonly employed exercise test consisting of repeated sprints on a cycle ergometer. Eight recreationally active young men completed a practice trial and three more trials at 3- to 6-day intervals. Each trial consisted of two bouts of 30-s maximal-effort cycling on an electromagnetically braked cycle ergometer; the bouts were separated by 4 min of rest. The typical (standard) errors of measurement for peak and mean power between trials 2 to 4 were 2.5 and 1.7% respectively for the first bout and 1.9 and 1.8% for the second bout. These errors are substantially less than those in previous reliability studies of single 30-s sprint tests, probably because of differences in quality of ergometer. The typical errors for the difference between bouts (i.e., fatigue) for peak power and mean power were 3.0 and 2.5%, respectively. Typical errors for the average of the two bouts were 1.6 and 1.2% for peak and mean power respectively, which are small enough to give adequate precision for moderate treatment effects in studies with modest sample sizes.

Psychometric evaluation of the body change inventory: an assessment instrument for adolescent boys and girls

A new instrument, the Body Change Inventory, was developed to provide an assessment of body change strategies that are used by both adolescent girls and boys. The novel aspect of this instrument is that it evaluates strategies to increase body size and increase muscle size, as well strategies to decrease body size. Independent samples of adolescent girls and boys aged between 11 and 17 years (N=1732) participated in four studies. The revised instrument consisted of three body change scales—Strategies to Decrease Body Size, Strategies to Increase Body Size, and Strategies to Increase Muscle Size. The studies demonstrated content validity, construct validity, internal consistency, and concurrent and discriminant validity for the new scales. The new scales provide a valuable addition in the literature for assessing three global body change strategies among adolescent girls and boys. They are needed in order to examine further the normative development of different kinds of body change strategies and how these may lead to behavioural problems such as disordered eating, exercise dependence, and steroid use.

Quality of life instruments for caregivers of patients with cancer : A review of their psychometric properties

The quality of life of caregivers of patients with cancer is an important construct given the substantial impact of caring on the physical, psychological, social, and financial well-being of caregivers. Moreover, caring for patients with cancer also affects family functioning and places burdens on caregivers. The reliability and validity of instruments used to assess the quality of life of caregivers of patients with cancer were reviewed to aid in the selection of the most appropriate measures For research and practice. MEDLINE (1980-2000) and c/NAHL (1982-2000) searches located relevant quality of life instruments using the keywords "cancer and quality of Iife" and "caregiver or spouse or partner." The search identified the following instruments: the Caregiver Quality of life Index-Cancer Scale, the 'Caregiver Quality of Life Index, the Quality of Life Tool, and the Quality of life Index-Cancer Version. Quality of life instruments developed specifically to measure the quality of life of caregivers of patients with cancer had the best psychometric properties. The Caregiver Quality of Life Index-Cancer Scale, in particular, met or exceeded minimum psychometric criteria for reliability and validity. The development of reliable and valid caregiver quality of life measures is an important factor in developing interventions to enhance quality of life of caregivers of patients with cancer.

Psychometric data on the mathematics self-efficacy scale

Descriptive measurement reliability and validity data are reported on scores from the Mathematics Self-Efficacy Scale (MSES) in two contexts—class and test—for a sample of 302 Australian high school students. Summated scores on the MSES correlated r = .74, and together these items yielded one component that explained 49% of the variance. MSES scores demonstrated internal reliability for both class and test (Cronbach alphas = .86 and .90). Statistically significant correlations between MSES scores, past mathematics grades, and Marsh’s Self-Description Questionnaire III (Math) were used to indicate the convergent validity of scores for the MSES. Discriminant validity of the MSES scores was suggested by the lack of significant correlations between the MSES measures and students’ desired English grades. Statistically significant correlations between MSES scores and students’ desired mathematics grades indicated concurrent validity of the MSES measures. A statistically significant difference between class and test scores supported Bandura’s prescription for context specificity in efficacy research.

Friction changes during compression tests using tool rotation technique

A compression test with rotating dies was employed to estimate the friction factor between aluminum samples and steel tooling during large plastic deformations. A cylindrical workpiece was compressed under dry and cold conditions. The magnitudes of torque and normal force were measured and
the average friction factor was calculated using the Coulomb friction law. It was found that under certain conditions the friction increased with increasing angular velocity of the die. This finding led to the conclusion that the choice of die rotation speed is important in interpreting the results from the twist compression test.

Robustness to non-normality of various tests for the one-sample location problem

This paper studies the effect of the normal distribution assumption on the power and size of the sign test, Wilcoxon's signed rank test and the t-test when used in one-sample location problems. Power functions for these tests under various skewness and kurtosis conditions are produced for several sample sizes from simulated data using the g-and-k distribution of MacGillivray and Cannon.

Psychometric properties of the quality of life questionnaire for children with CP

This paper describes the development and psychometric properties of a condition-specific quality of life instrument for children with cerebral palsy (CP QOL-Child). A sample of 205 primary caregivers of children with CP aged 4 to 12 years (mean 8y 5mo) and 53 children aged 9 to 12 years completed the CP QOL-Child. The children (112 males, 93 females) were sampled across Gross Motor Function Classification System (GMFCS) levels (Level I=18%, II=28%, III=14%, IV=11%, V=27%). Primary caregivers also completed other measures of child health (Child Health Questionnaire; CHQ), QOL (KIDSCREEN), and functioning (GMFCS). Internal consistency ranged from 0.74 to 0.92 for primary caregivers and from 0.80 to 0.90 for child self-report. For primary caregivers, 2-week test-retest reliability ranged from 0.76 to 0.89. The validity of the CP QOL is supported by the pattern of correlations between CP QOL-Child scales with the CHQ, KIDSCREEN, and GMFCS. Preliminary statistics suggest that the child self-report questionnaire has acceptable psychometric properties. The questionnaire can be freely accessed at http://www.deakin.edu.ac/hmnbs/chase/cerebralpalsy/cp_qol_home.php.

The factor structure of the strengths and difficulties questionnaire

The Strengths and Difficulties Questionnaire (SDQ) has been widely adopted as a measure of child and adolescent mental health and behavioral problems. However, despite an impressive number of studies demonstrating its psychometric properties, appropriate analyses of the instrument's underlying structure have not been reported. We conducted confirmatory factor analyses on matched data obtained from a large community sample of 7-17-year-olds, their parents, and their teachers. The analyses indicated that the reputed factor structure was not supported, with none of the subscales being unidimensional. Further, each informant group appeared to respond differently to the questionnaire. Considering the findings in relation to the stringency of the tests used to evaluate the SDQ, its utility, and previous research on its clinical validity, it is suggested that the SDQ be used with caution and in conjunction with other forms of assessment.