Statistical Assignment on HIV Patients
Statistical Assignment on HIV Patients
Reference: Haseli, N., Esmaeelzadeh, F., Ghahramani, F., Alimohamadi, Y., Hayati, R., &Mahboubi, M. (2014). Health-related quality of life and its related factors in HIV+ patients referred to Shiraz Behavioral Counseling Center, Iran in 2012. Medical Journal Of The Islamic Republic Of Iran, 281-6.
- Summary of the study
The study aims to evaluate the health-related quality of life and individual factors which influence QOF in HIV + patients. A cross-sectional study was carried out on 129 HIV+ patients who were selected by convenience sampling. Data collection was by the use of questionnaires based on demographic data and self or interviewer-administered questionnaires. The analysis was by chi test, ANOVA, T-test, and Schiff’s post hoc test. There was a significant difference in the quality of life of HIV + patients in terms of gender, marital status, employment status, and drug abuse history
- Identify the null hypothesis and alternative hypotheses
The null hypothesis is that there would be no significant difference in the quality of life of HIV+ affected patients with respect to several factors which affect the quality of life. The alternative hypothesis is that the overall quality of life and factors which affect the quality of life would be significantly different in HIV + patients.
- Address the type of data required and whether assumptions of the test were met.
Performing ANOVA requires either parametric (score or interval data) or non-parametric (ranking or ordering based) data. Categorical data analysis requires Chi test. The data required would be both involving categorical and continuous variables as such a study needs to take into account many areas of the life of the participants and the questionnaires used were able to generate the kind of data required.
The Dependent Variable (DV) here refers to the Quality of life of HIV + patients. It was measured in terms of the mean (average of all the data values) and the Standard Deviation of the values was noted. P values were obtained to determine the statistical significance of the results and whether the alternate hypothesis was valid.
The Independent Variable (IV) includes the set of demographic and QOL related variables which would or would not have impact on the QOL. A large number of these variables were studied by calculating mean scores and SD (standard deviations)
The data which were obtained included the kind of data required for ANOVA analysis (score/interval as well as categorical variables). Assumptions for ANOVA are that all the different groups analysed are just samples of the same population and that all treatments would have the same effect if the null hypothesis is true. The null hypothesis was found true for some cases but not all. Other assumptions are that the observations are independent, the distributions should be normal and the variance of data in groups are the same.
Comparisons of these variables between groups were carried out by ANOVA, chi square and t test to determine relationships between variables. The paper mentioned that Schiff’s post hoc test was used to determine significant levels of difference in QOL between the different variables studied.
Appendices: Guide for selecting inferential and descriptive. (2014). www.sagepub.com. Retrieved 18 April 2014, from http://www.sagepub.com/fitzgerald/study/materials/appendices/app_a.pdf
Levels of Measurement. (2014). Polsci.wvu.edu. Retrieved 18 April 2014, from http://www.polsci.wvu.edu/duval/ps601/Notes/Levels_of_Measure.html#Tools of Inference
Using Descriptive Statistics. (2014). Ncsu.edu. Retrieved 18 April 2014, from http://www.ncsu.edu/labwrite/res/gt/gt-stat-home.html
Section B Some of the results were analyzed using a T-test and others using an F test. Why?
Some of the results relating to Mean +/-SD quality of life in HIV+ patients as well the QOL for HIV positive patients in three different age groups in the above paper were analysed using t-test and p-value was determined. (Haseli et al 2014). The t-test refers to independent samples or paired samples t test. These tests are useful to perform when you are studying a limited number of variables (not more than two or three) or different levels of a variable. The results and scores obtained would be more relevant and significant. Here either different levels of a variable or effect of different variables on the QOL of same sample was studied. Hence t tests were used. Independent samples t tests can be used when you are comparing the effect on the dependent variable of study for two different samples using a single independent variable. It can also be used for studying different levels of an independent variable (two different levels). The t value would indicate the statistical significance as in how much the two samples differ or are similar in the test experience. A t value comparable to the scores of both the two individual scores for the different samples or the different IV levels would indicate that there is not much difference of effect on the two samples. An example of such problems would be when you are studying the effect of a drug on two different classes of patients belonging to two different age groups. The independent variable would be the effect of drug. The dependent variable would be the kind of outcome, which you would want to measure the effect by. The levels of the independent variable would be the different ages. A paired sample t- test would be useful when you want to try out the effect of different levels of the independent variable on the same sample. For e.g. effect of two drugs on the same sample. So, those problems with limited number of variables or levels of variables being studied are more appropriately studied with t tests. (Significance, 2014)
F- test or ANOVA was used for the comparison of means from different samples (sub groups such as divided on basis of age and sex) under different categories of independent demographic variables and for determining the overall QOL as an effect of factors such as marital status and drug abuse in different subgroups of HIV + patients. (Haseli et al 2014)
The F test is a measure used for ANOVA analyses which is used for comparison of the means from different samples and the results from which can be extrapolated to different populations. It predicts about the statistical significance of your result, as to how likely it could be, given that the two values compared were not really different. i.e the individual variances are not much different. (Background -Anova, 2014) (F-stat and significance, 2014)
ANOVA F test is more useful when there are more than two levels of a variable or many independent variables being studied. Using an F test is more useful as it will reduce the “error of inflation” which may occur if you perform multiple t tests for larger number of levels of variables or many variables. Hence, in those studies where you have many parameters being analysed, F test is more accurate.
Background – ANOVA. (2014). Afni.nimh.nih.gov. Retrieved 18 April 2014, from http://afni.nimh.nih.gov/pub/dist/HOWTO/howto/ht05_group/html/background_ANOVA.html
F-STAT AND SIGNIFICANCE. (2014). Tulane.edu. Retrieved 18 April 2014, from http://www.tulane.edu/~panda2/Analysis2/two-way/fstat_and_significance.htm
Significance. (2014). Usersfsu.edu. Retrieved 18 April 2014, from http://userwww.sfsu.edu/biol240/labs/lab_01variation/pages/1significance.html