Statistical analysis is a method used to process complicated data. Following are different types of statistical analysis.
1. General linear model
The General Linear Model (GLM) is a statistical method which is used in relating responses to the linear sequences of predictor variables including different types of dependent variables and error structures as specific cases. GLM states that most of the statistical analyses are used in social and applied research.
It is an extension of linear multiple regression for a single dependent variable. The one difference between multiple regression model and GLM is the number of dependent variables that can be analyzed. Within infinite number of experimental designs this model can be used in answering different research questions.
General linear model is the basic method for the Analysis of Variance (ANOVA), Analysis of Covariance (ANCOVA), t-test, f-test, regression analysis, and most of the multivariate techniques like canonical correlation, cluster analysis, discriminant function analysis, factor analysis, multidimensional scaling, and many more. Because of its generalization, this model is essential for social research pupil. GLM requires advanced statistical training, even though there is a deep understanding. GLM is used in testing about any theory and a dependent variable (DV) which is measured numerically (e.g., age, grade point average, income, height, etc). But categoric Dvs are not measured like sex, eye color etc.
Regression is a univariate general linear model. Univariate GLM is a method which is used in the Analysis of Variance for experiments having two or more factors. The chief panel demands for Dependant Variable, Covariants, Fixed effect factors, weighted least square and Random effect factors.
Multivariate GLM is a method used to perform the analysis of variance for experiments with more than one dependent variable. The main panel demands for fixed factors, weighted least square, covarieties, dependent variables and random effect factors.
2. Correlational statistical analysis
Correlation is a technique which looks for a relationship between two variables. A correlation test comprises of calculating the correlation coefficient from the two data sets and this coefficient is compared to an appropriate entry in correlation coefficient criterion numbers table. The criterion is selected based on the number of data pairs in the set. When the coefficient is higher or same as the selected criterion, it is said to be a significant relation between the two data sets.
The two popular correlation coefficients are pearson’s product moment correlation coefficient and spearman’s correlation coefficient.
To calculate the correlation coefficient of the original data opt for spearman’s correlation coefficient.
The correlation coefficient value may vary from minus one to plus one. When the correlation value is minus one, it indicates there is a negative relationship between the two variables. If the correlation value is plus one it indicates a positive relationship between the variables. If the value of correlation coefficient is zero, it indicates there is no relationship between the variables.
The confidence intervals about a true correlation of zero are determined by the standard error of correlation coefficient.
3. Chi-square analysis
Chi-square analysis is used to test the relation between two categorical variables. In this test when the null hypothesis is true, the sampling distribution of the test statistic is chi-squared distribution. Chi-square test is used in comparing observed data and the expected data to be obtained according to a specific method.
4. Sign test
Sign test is applied only if the two associated variables are different. It is comparison test on two samples. When observations are made for each pair then the higher observation is assigned as plus, poorer observation as minus and zero when there is no difference.
5. Wilcoxon rank sum test
It is used to test the deviation between two variables. It is a non-parametric method. It is used to test null hypothesis.