For the most widely used statistical methods, SPSS provides several ways of access to them. We can specify descriptive statistics analysis by selecting Analyze – Descriptive Statistics – Descriptives… or by selecting Analyze – Descriptive Statistics – Explore… Usually such different ways of selecting the same statistics have some peculiarities. For example, in the Descriptives… we can manually choose which descriptive statistics we want to assess, but this is not possible in option Explore… In the Explore… together with descriptive statistics we can assess the form of distribution by building graphs and calculating several criteria.
Let us start from Descriptives…:
1) Click the Analyze menu, then point to Descriptive Statistics, and select Descriptives… . The Descriptives dialog box is opened.
2) In Descriptives: Options dialog box select the variables for the analysis. In our example all four variable should be selected and after clicking the transfer arrow button the selected variables are moved to the Variable(s): list box.
3) Next click the Options… button:
Descriptives: Options dialog box opens:
4) Select the Mean, Std. deviation, Minimum, Maximum, S.E. mean, Kurtosis and Skewness check boxes.
5) Click the Continue button. This returns you to the Descriptives dialog box.
6) Click the OK button. An Output Viewer window opens and displays the statistics.
In the table with the results of descriptive statistics analysis we can see not only minimum, maximum, mean with its standard error, standard deviation, but also skewness and kurtosis, which give preliminary information about deviation of a distribution from the normal one: if standard errors of skewness and kurtosis are much smaller than their values (not of the same order), this indicates that distribution deviates from the normal.
As a first approximation, it may be useful to perform exploratory data analysis on an entire dataset to get a quick feeling for what the data looks like. However, if we are going to compare groups (subsets of the data) using a parametric test, we need to ensure that each of the groups being compared has a normal frequency distribution, so we have to perform exploratory analysis on all the groups separately.
A general way of splitting the file into subsets:
1) Click the Splitfile tab in the Tool Bar:
The Split File dialog box opens:
2) Select the variable “Species” from the left list box.
3) Select the Organize output by groups option.
4) Click the transfer arrow button to move the variable “Species” to the Groups Based on: list box.
5) Click the OK button.
For the descriptive statistics, the display of results by subgroups and also the selection of some more advanced statistics can be done by one more way:
1) Click the Analyze menu, point to Compare Means, and select Means… :
The Means dialog box is opened:
2) Select the dependent and independent variables. In our example all four MICs of oils are dependent variables and “Species” is independent, they should be selected, and after clicking the transfer arrow buttons for dependent and independent lists the selected variables are moved to the Dependent list or Independent list boxes, respectively:
3) Click the Options… button; the Means: Options dialog box opens:
4) The Mean, Number of Cases and Standard Deviation are already selected by default. Select also the Std. Error of Mean, Kurtosis, Std. Error of Kurtosis, Skewness and Std. Error of Skewnesstests; click the transfer arrow button and they will be moved to the Cell Statistics: list box:
5) Click the Continue button. This returns you to the Means dialog box.
6) Click the OK button. An Output Viewer window opens and displays the statistics:
Therefore, we not only specified descriptive statistics for analysis but also selected display of results for subsets of data, where subsets are defined by the values of the independent variable (“Species”).