The Technology in Training which was held a week ago by the Learning Technology Development Unit (UPTB) of the Faculty of Psychology UGM was resumed on Friday (12/3) with the same theme, namely “Data Processing with Jamovi: An Alternative to SPSS” and filled with Wahyu Jati Anggoro. , S.Psi., MA. This event was held at 13.00 WIB and opened with a discussion of descriptive statistics. “If we talk about descriptive statistics, yes, one of which is that we report the mean, median, standard deviation, minimum value, maximum value, and usually the report is displayed before we report the hypothesis test,” said Wahyu.
Then the discussion continues about data categorization using the method for classifying categorical data based on the mean plus 1 standard deviation unit or minus 1 standard deviation unit. The high category can use the formula for the mean value plus the value of one standard deviation, while the low category uses the formula for the mean minus one standard deviation. Before categorizing the data, researchers should first know the mean and standard deviation of the variables to be categorized.
Wahyu further explained that if the researcher uses SPSS, it will take two times to work, while categorizing the data using JAMOVI is enough to work once. Data categorization in JAMOVI can be done using create new transform then add record conditions. “So we will recondition the continuous data, how much the score range falls into a certain category,” said Wahyu. The next step if you want to find out how many numbers from each categorization can use the exploration menu and enter the variables that have been categorized.
After categorization, Wahyu repeated his explanation about Paired and Independent sample t-test. Paired Sample T-test is an analysis to test whether there is a difference between the dependent variable and the independent variable on the same subject and is commonly used in experimental research. Independent Sample T-test to test whether there is a difference between the dependent variable and the independent variable in 2 different groups. In this event, the topics discussed were looking for differences between data before and after training and data between men and women.
Furthermore, Wahyu explained the One-Way ANOVA to see the differences in each variable in terms of the origin of the department. “Now, to see whether there is a difference or not, we can check for those in the section on the significance level,” said Wahyu. If differences are found in data, further analysis can be carried out, namely the post-hoc test to determine the differences between categories in these variables. There are several further analysis options available in the post-hoc test section, including Tukey’s which is used when the variance is equal or the assumption of homogeneity of the variance is met and Games-Howell for the unequal variance.
Then, the discussion continues on the topic of regression. When using JAMOVI, the regression menu offers several options, one of which is Linear Regression. Before performing Linear Regression, there is an assumption used, namely colinearity to see if there is multicollinearity between variables. “There are two types of multicollinearity, namely VIF and tolerance. Some books show that the VIF is below 0.5 or 0.10. Then the tolerance is 0.2 or 0.3 ”, explained Wahyu.
At the end of the event, Wahyu emphasized that researchers should really understand, “at least the philosophycal analyzes behind every analysis that we click so, oh Anova, the philosophy is like this, the correlation is like this”. Although psychology is not a science that only studies statistics, it cannot be denied that psychology applies statistics in its scientific field. Therefore, at least people who are in the field of psychology still understand the philosophy, analysis, and interpretation of each research result. “More importantly, in psychology, numbers will not be meaningful if we cannot explain the psychological dynamics of each of the variables studied,” said Wahyu.