Kaiser-Guttman Rule Research Articles

Caregiver strain among patients of palliative care in Sri Lanka: validation of modified caregiver strain index - Sinhala version

BackgroundCare givers of Palliated patients are at risk of adverse physical, psychosocial and emotional sequelae in varied nature. Efficient and valid assessment tools facilitate early detection to take corrective measures. The Modified Caregiver Strain Index (MCSI), composed of domains associated with caregiver strain is a simple and brief tool that can be used in both clinical and field settings. This study aimed to adapt and validate this in order to cater effective palliative care services in Sri Lanka.MethodsAfter cross-cultural adaptation, 200 primary caregivers in 3 teaching hospitals were recruited. The internal consistency, item-total correlations, of the 13-item S-MCSI were performed. The criterion validity was assessed by Pearson correlation between the total scores of S-MCSI, the Karnofky Performance Scale and the Barthel index. Construct validity was determined by the principal component analysis keeping the Varimax with Keiser normalization as the rotation method. The Kaiser-Meyer-Olkin test (KMO) and Bartlett’s test of sphericity statistics were also performed to determine the adequacy of the sample and correlations between items, respectively. The number of factors was determined by the Scree plot, percentage of variance explained by each component and number of Eigen values over 01 (Kaiser-Guttman rule).ResultsThe total MCSI score ranged 0 to 26. The overall Cronbach’s alpha of the 13-item questionnaire was 0.80 while item-total corrections ranged 0.34 to 0.62, exception of one item (0.11). Inverse correlations were demonstrated in total scores of MCSI and Karnofky Performance Scale (r =- 0.32, p < 0.001) and Barthel index (r =-0.34, P < 0.001). A Kaiser-Meyer-Olkin value of 0.79 (p < 0.001) for Bartlett’s test indicated adequate sampling and nonlinearity of factors. The Scree plot showed a three-factor structure explaining 57% of the variation. Items regarding personal wellbeing of caregiver loaded together while the effects on the family loaded separately. Adjustment of personal concerns and family issues along with time alteration grouped as the third factor.ConclusionsThe study showed that the Sinhala version of MCSI has adequate psychometric properties and reliability to be used as a validated tool to estimate the caregiver burden within a short time period for any health care workers.

A Dichotomous Behavior of Guttman-Kaiser Criterion from Equi-Correlated Normal Population

We consider a p-dimensional, centered normal population such that all variables have a positive variance σ2 and any correlation coefficient between different variables is a given nonnegative constant ρ &lt; 1. Suppose that both the sample size n and population dimension p tend to infinity with p/n → c &gt; 0. We prove that the limiting spectral distribution of a sample correlation matrix is the Marcenko-Pastur distribution of index c and scale parameter 1 − ρ. By the limiting spectral distributions, we rigorously show the limiting behavior of widespread stopping rules Guttman-Kaiser criterion and cumulative-percentage-of-variation rule in PCA andEFA. As a result, we establish the following dichotomous behavior of Guttman-Kaiser criterion when both n and p are large, but p/n is small: (1) the criterion retains a small number of variables for ρ &gt; 0, as suggested by Kaiser, Humphreys, and Tucker [Kaiser, H. F. (1992). On Cliff’s formula, the Kaiser-Guttman rule and the number of factors. Percept. Mot. Ski. 74]; and (2) the criterion retains p/2 variables for ρ = 0, as in a simulation study [Yeomans, K. A. and Golder, P. A. (1982). The Guttman-Kaiser criterion as a predictor of the number of common factors. J. Royal Stat. Soc. Series D. 31(3)].

An Exploratory Step Toward Measuring the "Meaning of Life" in Patients with Tinnitus and in Cochlear Implant Users.

Many questionnaires attempt to quantify the "quality of life." However, we believe understanding the quality of life is complex, and many widely used questionnaires do not capture the broad range of factors that we believe are important. Many do not include questions about communicating. We developed a preliminary questionnaire designed to measure "The Meaning of Life" from a broader perspective. We reviewed other scales and sought input from individuals with disabilities and developed an initial 23-item questionnaire. As a first step, we sampled 116 adults with tinnitus and 196 with cochlear implants (CIs). Individuals who were participating in our CI or tinnitus research programs participated. To compare differences between the two participant groups, independent sample f-tests were completed for specific items on the questionnaire and for the total score. We compared age and gender differences across all participants using f-tests. Statistical significance was defined as p < 0.05. An exploratory factor analysis was conducted to examine the relationship among the questionnaire items using oblique rotation to produce correlated factors. Extracted factors with an eigenvalue >1.0 were retained according to the Kaiser-Guttman rule. Four factors were prominent in this initial sample, which we labeled (1) friendship and positive outlook, (2) physical health, (3) hearing and mental health, and (4) satisfaction with life. Participants with tinnitus reported more trouble sleeping than participants with CI, whereas both groups had lower scores on hearing. Older patients reported more difficulty with remembering things but were more satisfied with their financial situation. Female participants reportedly had more hobbies and were more satisfied with their sex lives than male participants. This exploratory study intended to take a broader look at quality of life scales. Further work is needed with a larger sample including younger and older participants with and without disabilities.

Learning effects of different training models for border molding from the perspective of dental students

BackgroundThe aim of this study was to investigate the effects of different practical training models on the comprehension and evaluation of practical training among dental students.MethodsThe study subjects were all sixth-year dental students at our institute, and the study took place over three consecutive years (n = 58, 63, and 65, respectively). In practical training, all students learned border molding, and practical models were modified each year from plaster models to silicone models and then to silicone models mounted in mannequins. Immediately after completing clinical training, all students were asked to complete questionnaires consisting of 21 items regarding their overall practical training and their clinical comprehension of border molding. All items were rated on a five-point Likert scale, and in order to reduce the large number of interrelated questions, exploratory factor analysis was carried out using maximum likelihood estimation with promax rotation (κ = 4) and Kaiser normalization. The number of factors was chosen using the Kaiser-Guttman rule, which states that the eigenvalue should be larger than 1, and the scree plot criteria. Items that scored less than 0.25 in communality and exhibited factor loading greater than 0.35 for more than one item were excluded. The defined factors were analyzed for the plaster models, the silicone models alone, and the silicone models with mannequins using the Kruskal-Wallis test and follow-up tests using Bonferroni-corrected Mann-Whitney U tests. The significance level was set at p < 0.05.ResultsExploratory factor analysis identified the following three factors: “knowledge of border molding”; “contents of practical training”; and “personal learning attitude”. The students who used silicone models and mannequins gave significantly better evaluations on the “knowledge of border molding” (p < 0.001, both) and “contents of practical training” (p = 0.046, p < 0.001, respectively) subscales than those who used plaster models. No significant differences were observed between those who used silicone models and those who used mannequins. Moreover, no significant differences were found on the “personal learning attitude” subscale among students for any model.ConclusionsThe change in practical training models from plaster to silicone improved student evaluations of border molding training.

Exploratory graph analysis: A new approach for estimating the number of dimensions in psychological research.

The estimation of the correct number of dimensions is a long-standing problem in psychometrics. Several methods have been proposed, such as parallel analysis (PA), Kaiser-Guttman’s eigenvalue-greater-than-one rule, multiple average partial procedure (MAP), the maximum-likelihood approaches that use fit indexes as BIC and EBIC and the less used and studied approach called very simple structure (VSS). In the present paper a new approach to estimate the number of dimensions will be introduced and compared via simulation to the traditional techniques pointed above. The approach proposed in the current paper is called exploratory graph analysis (EGA), since it is based on the graphical lasso with the regularization parameter specified using EBIC. The number of dimensions is verified using the walktrap, a random walk algorithm used to identify communities in networks. In total, 32,000 data sets were simulated to fit known factor structures, with the data sets varying across different criteria: number of factors (2 and 4), number of items (5 and 10), sample size (100, 500, 1000 and 5000) and correlation between factors (orthogonal, .20, .50 and .70), resulting in 64 different conditions. For each condition, 500 data sets were simulated using lavaan. The result shows that the EGA performs comparable to parallel analysis, EBIC, eBIC and to Kaiser-Guttman rule in a number of situations, especially when the number of factors was two. However, EGA was the only technique able to correctly estimate the number of dimensions in the four-factor structure when the correlation between factors were .7, showing an accuracy of 100% for a sample size of 5,000 observations. Finally, the EGA was used to estimate the number of factors in a real dataset, in order to compare its performance with the other six techniques tested in the simulation study.

Construction and validation of a dimensional scale exploring mood disorders: MAThyS (Multidimensional Assessment of Thymic States)

BackgroundThe boundaries between mood states in bipolar disorders are not clear when they are associated with mixed characteristics. This leads to some confusion to define appropriate therapeutic strategies. A dimensional approach might help to better define bipolar moods states and more specifically those with mixed features.Therefore, we proposed a new tool based on a dimensional approach, built with a priori five sub-scales and focus on emotional reactivity rather than exclusively on mood tonality. This study was designed to validate this MAThyS Scale (Multidimensional Assessment of Thymic States).MethodsOne hundred and ninety six subjects were included: 44 controls and 152 bipolar patients in various states: euthymic, manic or depressed. The MAThyS is a visual analogic scale consisting of 20 items. These items corresponded to five quantitative dimensions ranging from inhibition to excitation: emotional reactivity, thought processes, psychomotor function, motivation and sensory perception. They were selected as they represent clinically relevant quantitative traits.ResultsConfirmatory analyses demonstrated a good validity for this scale, and a good internal consistency (Cronbach's alpha coefficient = 0.95). The MathyS scale is moderately correlated of both the MADRS scale (depressive score; r = -0.45) and the MAS scale (manic score; r = 0.56).When considering the Kaiser-Guttman rule and the scree plot, our model of 5 factors seems to be valid. The four first factors have an eigenvalue greater than 1.0 and the eigenvalue of the factor five is 0.97. In the scree plot, the "elbow", or the point at which the curve bends, indicates 5 factors to extract. This 5 factors structure explains 68 per cent of variance.ConclusionThe characterisation of bipolar mood states based on a global score assessing inhibition/activation process (total score of the MATHyS) associated with descriptive analysis on sub-scores such as emotional reactivity (rather than the classical opposition euphoria/sadness) can be useful to better understand the broad spectrum of mixed states.

Teacher’s Corner: Toward a Coherent View of Reliability in Test Theory

Reliability of test scores, as estimated through measures of internal consistency, has been characterized mathematically in many ways that appear, on the surface at least, to be very dissimilar to one another. In this essay we provide a general mathematical framework that specializes to four different reliability coefficients. Through consideration of this general framework it becomes easier to convey to students both the individual character of the different formulations of reliability and the extent of their underlying similarity. In addition to providing a coherent view of reliability, the unified formula is also found to be a convenient vehicle for introducing more specialized topics, such as the Kaiser-Guttman rule.

Toward a Coherent View of Reliability in Test Theory

Reliability of test scores, as estimated through measures of internal consistency, has been characterized mathematically in many ways that appear, on the surface at least, to be very dissimilar to one another. In this essay we provide a general mathematical framework that specializes to four different reliability coefficients. Through consideration of this general framework it becomes easier to convey to students both the individual character of the different formulations of reliability and the extent of their underlying similarity. In addition to providing a coherent view of reliability, the unified formula is also found to be a convenient vehicle for introducing more specialized topics, such as the Kaiser-Guttman rule.

On Reliability and Number of Principal Components: Joinder with Cliff and Kaiser

This note summarizes my remarks on the application of reliability of the principal component and the eigenvalue-greater-than-1 rule for determining the number of factors in principal component analysis of a correlation matrix. Due to the unpredictability and uselessness of the reliability approach and the Kaiser-Guttman rule, research workers are encouraged to use other methods such as the scree test.

Derivations of the Reliability of Components

This paper attempts to resolve the disagreement between Cliff (1988) and Kaiser (e.g., 1991) concerning the reliability of components as estimated by the Kaiser-Guttman rule. It is suggested that the differences result from starting with a definitional formula for reliability in the former as opposed to substituting in a concluding formula in the latter. The length of time during which the formula was not challenged is remarked.

On Cliff's Formula, the Kaiser-Guttman Rule, and the Number of Factors

Cliff (1988) has presented a formula for the reliability of a principal component which is different from my long-known formula (Kaiser, 1957, 1991) for coefficient alpha of a principal component. Cliff claims that his approach is “correct” and mine “is the result of a misapplication of the formula for internal consistency reliability” Actually, both developments are correct but are based on different premises: Cliff considers measurement error within—but not between—attributes, while I consider measurement error between—but not within—attributes. The application of my formula to the knotty problem of the “number of factors”—the Kaiser-Guttman Rule—appears often to give the “right” result, when “right” means agreement with the subjective judgment of factor-analytic grandmasters. But when it fails it is approximately equally likely to overfactor as to underfactor. Cliff's formula, on the other hand, when used to establish the number of factors, almost invariably overfactors and, in the limit, as the within-attribute reliabilities all approach one (as with, say, physical attributes), nonsensically will declare all principal components perfectly reliable no matter how small their associated eigenvalues, yielding an absurd answer if used to establish the number of factors.

Coefficients Alpha for Components

The formula for coefficient alpha for a component is developed. Kaiser's special formula for coefficient alpha for a principal component and the Kaiser-Guttman Rule for the "number of components" are mentioned. A numerical example is added. It is conjectured that the sum of coefficients alpha is invariant under orthogonal rotation.

Coefficient Alpha for a Principal Component and the Kaiser-Guttman Rule

The formula for coefficient alpha for a principal component, first derived in 1957 by Kaiser, is developed. Its use in the Kaiser-Guttman Rule for the number of components is discussed, both in theory and in practice, with Hotelling's (1933) original correlation matrix.