Abstract
Assessment is an important phase of the teaching and learning processes. Assessment generates a big amount of data, which in the case of summative assessment is in the form of students' grades. In e-assessment application, these grades are easily collected and stored. The analysis of the grades is very important to directly evaluate e-assessment and thus indirectly evaluate the teaching and learning processes. This chapter presents a practical example of analysis of grades obtained during an e-assessment process implementation using statistical methods. Important tips on how to correctly use these statistical methods are presented throughout the chapter. The analysis concerns seven years, and a positive evolution of the grades is verified.
TopMethod
The evolution of the final grades (the scale of the grades ranges from 1 to 20) of the college students in two curricular units throughout seven years was analyzed, considering the average of grades and the proportion of positive grades, in three clearly distinct periods of implementation of an e-assessment strategy.
All the data referring to the grades of the students presented in this chapter were collected from the database of the Institution, and with the proper authorization of the Dean of the school.
After being collected, the data were later processed, as they had some coded information that needed to be corrected, for example, students with grade “88” were students who had in the meantime given up. These students were removed from the database. Another situation that was corrected was that many students had more than one grade in the same curricular unit and the same year, because the database contained the grades of the several exams that the student had carried out during that year (continuous assessment, second examination period, etc.). The repetitions were eliminated leaving only the higher grade, since this is the grade that will be assigned to the student.
In order to analyze and interpret the data, we made use of descriptive statistics and statistical inference, using MS Excel™ as the main working tool. Within the scope of the descriptive statistics, tables and graphs were constructed, and the calculation of some localization and dispersion measurements were performed, which essentially summarize and describe the data. In the scope of the statistical inference, among other tools, we used several hypothesis tests, in particular applying the variance analysis (one-way ANOVA), which allowed us to draw conclusions on the data.
Key Terms in this Chapter
Add-In/Plug-In: An application that gives a new functionality or a new feature to the software.
Statistics: Science which analyses and draws conclusions from data. It can be applied in most areas of human knowledge.
Statistical Data Analysis: Data analysis by statistical methods, which ranges from the collection of data, the measures using data, to the interpretation of data.
ANOVA (Analysis of Variance): One of the most used statistical methods. In simple terms, this method tests groups to see if they have some differences among them.
Statistical Tests/Methods: A huge set of statistical tools that allows to make inferences (probabilistic assertions about some population, based on the analysis of data from a sample of that population).
Size of the Effect: This is a descriptive statistic. In some data, it is a complement to some statistical tests. In other analysis this is a very important measure to quantify the effect, evaluate/quantify the difference between two groups. Some examples of size of effect measures are Cohen d, RMSSE, and Omega Square.