The analysis of quantitative data is a deductive process. Depending on the nature of the assessment, faculty members will use either descriptive or inferential statistics to analyze the data.
Descriptive statistics typically fall into four basic categories: (a) frequency distributions, (b) measures of central tendency, (c) measures of dispersion, and (d) measures of relationship. Frequency distributions appear as tables with the number of responses in each category, arranged in order. They may also be displayed graphically as bar charts, histograms, or pie charts. Measures of central tendency include the mean (the arithmetical average of scores), the median (the middle score), and the mode (the most commonly occurring score). The selection of the appropriate measure of central tendency depends on the type of data and whether they are normally distributed. Measures of dispersion explain the range or spread of scores within the group. This may be expressed as a standard deviation or an inter-quartile range, again depending on the type of data. Measures of relationship are typically expressed as correlation coefficients show the direction and strength of two the relationship variables.
Some questions require inferential statistical analysis. The most common statistical analytical techniques for educational research are independent samples t-tests and chi-square tests of independence, with the choice of test depending on the type of data collected. These and other statistical tests can easily be conducted using statistical software such as SPSS, which is available for faculty and staff to download from the BSU IT department. SPSS is also available in several computer labs on campus.
Last Modified: October 19, 2012