Solvella

Mini guide 3 steps · 2 min

Descriptive Statistics: summarise numbers in one go

Get the average, middle value, spread, smallest and largest for one or many columns at once. Example: test scores for three groups of 10 students.

Follow alongScores for Group A, B and C in A1:C11.
Excel (.xlsx)CSV

  1. Open Data Analysis and find the tool

    Click ExtensionsSolvellaData Analysis.... 1 Type desc in the search box, then click Descriptive Statistics.

    The search box at the top of the Data Analysis tool list
    Type a few letters to find a tool.Enlarge

  2. Fill in the form and press Run

    • 1 Input range: A1:C11, all three groups with their headings. Type it, or select the cells and click the small grid button in the box. It shows as $A$1:$C$11.
    • 2 Grouped by: Columns, because each group is a column.
    • 3 Tick Labels in first row and Summary statistics. Leave the confidence level at 95.
    • 4 Press Run. The results go on a new tab.
    The Descriptive Statistics form: Input range $A$1:$C$11, Grouped by Columns, Labels in first row and Summary statistics ticked, and the Run button
    The Descriptive Statistics form, filled in.Enlarge

  3. Read the results

    You get one block of two columns per group. What the main lines mean, using Group A:

    LineGroup AIn plain English
    Mean81.6The average score.
    Median81The middle score when they're sorted.
    Mode#N/AThe most common score. #N/A means no score appears twice.
    Standard Deviation8.13How far a typical score is from the average.
    Range26Highest minus lowest: 94 − 68.
    Count10How many scores there are.
    Confidence Level (95.0%)5.81The true average is very likely within 81.6 ± 5.81.
    Skewness, Kurtosis−0.08, −0.64The shape of the spread. Close to 0 means fairly even on both sides.
    Descriptive Statistics output for Group A: Mean 81.6, Standard Error, Median 81, Mode #N/A, Standard Deviation 8.13, down to Confidence Level(95.0%) 5.81
    Group A's results. Groups B and C sit to the right, laid out the same way.Enlarge
Next: to check whether the groups really differ, try t-Test or Anova: Single Factor from the same list. All Data Analysis tools