Analyses

Descriptives

Summarise one or more columns with counts, means, and frequencies.

Edit on GitHub

When to use it

Use descriptives when you want to see what a variable looks like before you test anything. They answer “what is typical, and how spread out are the scores?” For example: in a class of 24, what is the average exam score, and how many students sat each teaching method?

Assumptions

Descriptive statistics do not assume a normal distribution or equal groups. Tensr does not run a normality test as part of this procedure.

Each statistic uses the non-missing values in that column. A row can count toward the mean of one column and be skipped for another. The report also states how many rows are complete on every selected column.

Running it in Tensr

Analyze → Descriptive Statistics → Descriptives. In chat, ask for descriptives, for example “Summarise score and method.”
Put every column you want summarised into the column list. Leave it empty to include every column.
Tick the statistics you want. The dialog starts with mean, median, standard deviation, range, minimum, and maximum switched on. Variance, skewness, kurtosis, and the standard error of the mean start off.

Options

These names are the fields Tensr sends. The dialog labels the last three under Report output.

Prop

Type

Reading the output

Ninety-six students. score is the exam, hours is time studied, and method is lecture, online, or workshop. The request asks for the mean, median, standard deviation, range, minimum, and maximum.

Statisticscorehoursmethod
mean71.3954.984—
50%70.154.88—
std8.3081.222—
range41.96.52—
min49.42.14—
max91.38.66—

Summary statistics for 3 column(s); dataset N = 96. Summary statistics for 3 column(s); dataset N = 96. Metrics: Dataset rows = 96; Listwise-complete rows = 96; Columns = 3.

Reporting (APA 7)

Summary statistics for 3 column(s); dataset N = 96. Report the estimate in the table. This procedure is not summarised by one p-value.

Coming from SPSS

The path stored for this procedure is Analyze → Descriptive Statistics → Frequencies. Tensr’s own menu label is Descriptives. SPSS Descriptives (the other menu item) does not print a frequency table.

One-way ANOVA and the independent-samples t-test start from these same means. A histogram with a strange shape is a reason to look at Shapiro–Wilk before a parametric test.