BackgroundDifferent study designs and population size may require different sample size for logistic regression. This study aims to propose sample size guidelines for logistic regression based on observational studies with large population.MethodsWe estimated the minimum sample size required based on evaluation from real clinical data to evaluate the accuracy between statistics derived and the actual parameters. Nagelkerke r-squared and coefficients derived were compared with their respective parameters.ResultsWith a minimum sample size of 500, results showed that the differences between the sample estimates and the population was sufficiently small. Based on an audit from a medium size of population, the differences were within ± 0.5 for coefficients and ± 0.02 for Nagelkerke r-squared. Meanwhile for large population, the differences are within ± 1.0 for coefficients and ± 0.02 for Nagelkerke r-squared.ConclusionsFor observational studies with large population size that involve logistic regression in the analysis, taking a minimum sample size of 500 is necessary to derive the statistics that represent the parameters. The other recommended rules of thumb are EPV of 50 and formula; n = 100 + 50i where i refers to number of independent variables in the final model.

Cancer survivors in this middle-income setting have persistently impaired HRQoL and high levels of psychological distress. Development of a holistic cancer survivorship program addressing wider aspects of well-being is urgently needed in our settings.

The Maslach Burnout Inventory (MBI) has been widely used in research for more than 2 decades and is recognized as the leading measure of burnout. Malaysia is a multi-ethnic country and English is regarded as the second language. Therefore, it is essential to have a culturally acceptable translated Malay version of MBI, which can easily be understood by Malaysians, in order to study the burnout level in our population. Hence, the objectives of this study are to translate, cross-culturally adapt and validate specifically the Malay versions of the Maslach Burnout InventoryHealth Services Survey (MBI-HSS), Maslach Burnout Inventory -Educators Survey (MBI-ES) and Maslach Burnout Inventory -General Survey (MBI-GS) in Malaysia. Intraclass correlation was used to examine the test-retest reliability of the Malay versions of the MBI-HSS, MBI-ES and MBI-GS, while Cronbach"s alpha was used to assess the internal consistency of the subscales and the instruments as a whole. Exploratory factor analysis involving the principle component analysis extraction and varimax rotation were used to investigate the construct validity of the instruments. Overall, high intraclass correlation and Cronbach"s alpha values were achieved in the Malay versions of the MBI-HSS, MBI-ES and MBI-GS. The eigenvalue in factor analysis revealed all items in the Malay versions of the MBI-HSS, MBI-ES and MBI-GS can be grouped into 3 components, which were very similar to the original English versions. In conclusion, the findings from this study had demonstrated the Malay versions of the MBI-HSS, MBI-ES and MBI-GS were valid and appropriate to be used in Malaysia.

No abstract

Background: MLR and ANCOVA are common statistical techniques and are used for both experimental and non-experimental studies. However, both types of study designs may require different basis of sample size requirement. Therefore, this study aims to proposed sample size guidelines for MLR and ANCOVA for both experimental and non-experimental studies. Methods: We estimated the minimum sample sizes required for MLR and ANCOVA by using Power and Sample Size software (PASS) based on the pre-specified values of alpha, power and effect size (R2). In addition, we also performed validation of the estimates using a real clinical data to evaluate how close the approximations of selected statistics which were derived from the samples were to the actual parameters in the targeted populations. All the coefficients, effect sizes and r-squared obtained from the sample were then compared with their respective parameters in the population. Results: Small minimum sample sizes required for performing both MLR and ANCOVA when r-squared is used as the effect size. However, the validation results based on an evaluation from a real-life dataset suggest that a minimum sample size of 300 or more is necessary to generate a close approximation of estimates with the parameters in the population. Conclusions: We proposed sample size calculation when r-squared is used as an effect size is more suitable for experimental studies. However, taking a larger sample size such as 300 or more is necessary for clinical survey that is conducted in a non-experimental manner.