9 services

Descriptive Statistics & Exploratory Data Analysis in SPSS

Every quantitative study begins here. We compute frequencies, central tendency, variability, distribution shape, normality tests and cross-tabulations in SPSS — then deliver APA-formatted tables and charts ready for your Chapter 4.

9

Analysis Types

APA 7

Formatted Output

48hr

Typical Turnaround

Free

Initial Consultation

What Are Descriptive Statistics?

Descriptive statistics are the numerical summaries and visual displays that characterise a dataset. They answer the most fundamental questions about your data: What is the typical value? How spread out are the scores? Is the distribution symmetric or skewed? Are there any extreme values?

Unlike inferential statistics (t-tests, ANOVA, regression), descriptive statistics do not generalise beyond your sample or test hypotheses. Their purpose is to summarise, organise and present the data so that patterns become visible and interpretable. They are the foundation that every subsequent analysis rests on.

The Four Pillars of Descriptive Statistics

1. Measures of Central Tendency

The "typical" value in a distribution.

  • Mean — arithmetic average; best for symmetric, continuous data
  • Median — middle value; robust to outliers and skew
  • Mode — most frequent value; the only option for nominal data
  • Trimmed mean — mean after removing extreme percentiles

2. Measures of Variability

How spread out the scores are.

  • Range — maximum minus minimum; crude but simple
  • Interquartile range (IQR) — spread of the middle 50%
  • Variance (σ²) — average squared deviation from the mean
  • Standard deviation (SD) — square root of variance; same units as the mean

3. Measures of Distribution Shape

Symmetry and tail weight.

  • Skewness — negative = left tail, positive = right tail, 0 = symmetric
  • Kurtosis — excess kurtosis > 0 = heavy tails (leptokurtic), < 0 = light tails (platykurtic)
  • Rule of thumb — absolute skewness < 2 and absolute kurtosis < 7 is often considered acceptable for parametric tests

4. Frequency Distributions

How often each value or category occurs.

  • Frequency — raw count per category or value
  • Percent — proportion of total including missing
  • Valid percent — proportion excluding missing values
  • Cumulative percent — running total of valid percentages

Our 9 Descriptive Statistics Services

Each service below includes assumption verification, full SPSS output, plain-language interpretation and APA-formatted tables. Click any service to learn more or request a quote.

Frequency Analysis

Frequency distributions, percentages, valid percentages and cumulative percentages for categorical and ordinal variables.

Descriptive Statistics

Mean, median, mode, standard deviation, variance, skewness, kurtosis and range computed and interpreted.

Cross-Tabulation

Two-way and multi-way crosstabs with row, column and total percentages for categorical comparisons.

Exploratory Data Analysis

Distribution analysis, outlier detection and descriptive summaries using the Explore procedure.

Normality Assessment

Shapiro-Wilk, Kolmogorov-Smirnov, Q-Q plots, P-P plots and histogram assessment for distributional assumptions.

Distribution Analysis

Normal, binomial and Poisson distribution fitting and visual assessment.

Ratio Analysis

Ratio statistics and group ratios for proportional comparisons across cases.

Multiple-Response Analysis

Multiple-response frequencies and crosstabs for survey questions where respondents select more than one answer.

Graphical Analysis

Histograms, bar charts, pie charts, boxplots, line charts and scatterplots for publication-ready visuals.

How to Run Descriptive Statistics in SPSS — Step by Step

Below is the standard workflow our statisticians follow for every descriptive analysis. These steps apply to SPSS versions 25–29 and SPSS Subscription.

  1. 1

    Open and Inspect Your Dataset

    Open your .sav file or import data via File → Open → Data (for .sav) or File → Import Data (for Excel/CSV). Switch to Variable View and confirm each variable has the correct:

    • Name — short, meaningful, no spaces
    • Type — Numeric for numbers, String for text
    • Measure — Scale (continuous), Ordinal or Nominal
    • Values — value labels for coded variables (e.g. 1 = Male, 2 = Female)
    • Missing — user-defined missing values (e.g. 99 = Not applicable)
  2. 2

    Run Frequencies for Categorical Variables

    Analyze → Descriptive Statistics → Frequencies

    Move all nominal and ordinal variables into the Variable(s) box. Click Statistics to request mode, median and percentiles. Click Charts to select bar charts. Click OK. SPSS produces a frequency table for each variable showing frequency, percent, valid percent and cumulative percent.

  3. 3

    Run Descriptives for Continuous Variables

    Analyze → Descriptive Statistics → Descriptives

    Move all scale (continuous) variables into the Variable(s) box. Click Options and select: Mean, Std. deviation, Minimum, Maximum, Skewness, Kurtosis. Optionally check "Save standardized values as variables" to create z-score columns for outlier detection. Click Continue, then OK.

  4. 4

    Run Explore for Normality and Outliers

    Analyze → Descriptive Statistics → Explore

    Move continuous variables to Dependent List. Optionally move a grouping variable to Factor List to get separate statistics per group. Click Plots: check "Normality plots with tests" (produces Shapiro-Wilk and Kolmogorov-Smirnov tests plus Q-Q and P-P plots) and "Histogram". Click Continue, then OK.

    The Explore output includes a Case Processing Summary, Descriptives table (mean, 95% CI, 5% trimmed mean, median, variance, SD, minimum, maximum, range, IQR, skewness, kurtosis), Tests of Normality table, histograms, stem-and-leaf plots, boxplots and Q-Q plots.

  5. 5

    Run Cross-Tabulation

    Analyze → Descriptive Statistics → Crosstabs

    Place one categorical variable in Row(s) and another in Column(s). Click Cells: check Row, Column and Total percentages plus Expected counts. Click Statistics: check Chi-square, Phi and Cramér's V. Click Continue, then OK.

  6. 6

    Create Publication-Ready Charts

    Graphs → Chart Builder (or Graphs → Legacy Dialogs)

    For histograms: drag a scale variable to the x-axis in Chart Builder, select "Histogram", optionally overlay a normal curve. For bar charts: drag a categorical variable to the x-axis and select "Simple Bar". Double-click any chart in the Output Viewer to open the Chart Editor for formatting (fonts, colours, axis labels, reference lines).

  7. 7

    Interpret and Report Results

    Review each output table in the SPSS Output Viewer. Build your APA descriptive statistics table (see the APA Reporting section below). Note any variables with significant normality violations, high skewness/kurtosis or outliers — these inform your choice of parametric vs. nonparametric tests.

SPSS Menu Paths — Quick Reference

Bookmark this table. Every descriptive procedure in SPSS, with its exact menu path and corresponding syntax command.

Procedure Menu Path Syntax Command
Frequencies Analyze → Descriptive Statistics → Frequencies FREQUENCIES
Descriptives Analyze → Descriptive Statistics → Descriptives DESCRIPTIVES
Explore Analyze → Descriptive Statistics → Explore EXAMINE
Cross-tabulation Analyze → Descriptive Statistics → Crosstabs CROSSTABS
Ratio Statistics Analyze → Descriptive Statistics → Ratio RATIO STATISTICS
Multiple Response Analyze → Multiple Response → Define Variable Sets / Frequencies / Crosstabs MULT RESPONSE
Chart Builder Graphs → Chart Builder GGRAPH
Histogram Graphs → Legacy Dialogs → Histogram GRAPH /HISTOGRAM
Boxplot Graphs → Legacy Dialogs → Boxplot GRAPH /BAR or EXAMINE
P-P / Q-Q Plots Analyze → Descriptive Statistics → P-P Plots / Q-Q Plots PPLOT / EXAMINE

Which Descriptive Statistic for Which Variable?

The right descriptive statistic depends on your variable's measurement level. Use this decision table.

Measurement Level Examples Central Tendency Variability Charts
Nominal Gender, ethnicity, marital status, diagnosis Mode Frequency count, percentage Bar chart, pie chart
Ordinal Likert scales, education level, satisfaction rating Median, mode IQR, range, frequency Bar chart, boxplot
Interval Temperature (°C), IQ scores, year Mean, median SD, variance, range, IQR Histogram, boxplot, Q-Q plot
Ratio Age, income, weight, test score (0–100) Mean, median SD, variance, range, IQR, CV Histogram, boxplot, scatterplot

Tip: If a continuous variable is severely skewed (|skewness| > 1), report both the mean and median. The median is more representative of the typical value when the distribution is asymmetric.

Normality Testing in SPSS

Normality is a key assumption for parametric tests. Here is exactly how to assess it and what to do when it is violated.

Statistical Tests for Normality

Test Best For Null Hypothesis Decision Rule
Shapiro-Wilk n < 2,000 (preferred) Data are normally distributed p > .05 → assume normality
Kolmogorov-Smirnov n ≥ 2,000 Data follow specified distribution p > .05 → assume normality

Visual Methods for Normality

Histogram with Normal Curve

Overlay a normal curve on the histogram. If the bars roughly follow the bell shape, normality is reasonable. Look for obvious bimodality, floor/ceiling effects or long tails.

Q-Q Plot (Normal Probability Plot)

Data points should fall approximately along the diagonal reference line. Systematic S-curves indicate skewness; inverted-S patterns indicate kurtosis violations.

P-P Plot (Normal Probability Plot)

Similar to Q-Q but plots cumulative proportions. Points above the line = positive skew; below = negative skew. More sensitive to deviations in the centre of the distribution.

Boxplot

Assess symmetry (median line centred in box?), spread (box and whisker length) and outliers (circles and asterisks beyond whiskers). Useful for comparing distributions across groups.

When Normality Is Violated

If your data are not normally distributed, you have several options depending on severity:

  1. Mild violation (|skewness| < 1) — Parametric tests are generally robust. Proceed but report the violation.
  2. Moderate violation — Transform the data (log, square root, inverse) and re-check.
  3. Severe violation or small n — Use nonparametric alternatives (Mann-Whitney U, Kruskal-Wallis, Spearman).
  4. Large sample (n > 30 per group) — Central Limit Theorem applies; means are approximately normal even if individual scores are not.

How to Report Descriptive Statistics in APA Format

APA 7th edition has specific rules for presenting descriptive statistics. Here are the templates we follow for every deliverable.

APA Table Template — Continuous Variables

Variable N M SD Min Max Skew Kurt
Job satisfaction 215 3.42 0.87 1.00 5.00 −0.31 −0.18
Work experience (years) 215 8.76 4.52 0.50 28.00 1.12 0.84
Organisational commitment 210 3.89 0.73 1.33 5.00 −0.47 0.21

Note. M = mean; SD = standard deviation; Skew = skewness; Kurt = kurtosis.

APA Narrative Examples

Continuous variable

"Participants reported moderate levels of job satisfaction (M = 3.42, SD = 0.87, N = 215). Skewness (−0.31) and kurtosis (−0.18) values were within acceptable limits, suggesting an approximately normal distribution."

Categorical variable

"The majority of participants were female (n = 132, 61.4%), followed by male (n = 78, 36.3%) and non-binary (n = 5, 2.3%)."

Normality test

"The Shapiro-Wilk test indicated that work experience scores were significantly non-normal, W(215) = .94, p < .001. Visual inspection of the histogram confirmed a positive skew (skewness = 1.12). A log transformation was applied before parametric analysis."

APA Formatting Rules for Descriptive Tables

  • No vertical lines. Only three horizontal rules: top, header separator, bottom.
  • Use italics for statistical symbols: M, SD, N, n, W, p.
  • Report means and standard deviations to two decimal places.
  • Report skewness and kurtosis to two decimal places.
  • Use a minus sign (−), not a hyphen (-), for negative values.
  • Report exact p-values to three decimal places (e.g. p = .042), except when p < .001.
  • Number every table and give it a descriptive title in italics above the table.
  • Add a Note below the table defining abbreviations.

Worked Example: Employee Satisfaction Survey

A researcher surveys 215 employees on job satisfaction (5-point Likert), organisational commitment (5-point Likert), work experience (years) and department (categorical: HR, Finance, IT, Operations). Here is the complete descriptive analysis workflow.

Step 1: Sample Demographics

Run Frequencies on Gender and Department. Results show 61.4% female, 36.3% male, 2.3% non-binary. Department distribution: IT (28.4%), Operations (26.0%), Finance (24.2%), HR (21.4%). No missing data on demographics.

Step 2: Continuous Variable Summaries

Run Descriptives with mean, SD, min, max, skewness, kurtosis. Job satisfaction: M = 3.42, SD = 0.87, skew = −0.31, kurtosis = −0.18. Work experience: M = 8.76, SD = 4.52, skew = 1.12, kurtosis = 0.84. Organisational commitment: M = 3.89, SD = 0.73, skew = −0.47, kurtosis = 0.21.

Step 3: Normality Assessment

Run Explore with normality plots. Shapiro-Wilk results: job satisfaction W(215) = .98, p = .112 (normal); work experience W(215) = .94, p < .001 (non-normal); commitment W(210) = .97, p = .039 (borderline). Q-Q plots confirm work experience has a right-skewed distribution. Decision: log-transform work experience for parametric tests; other variables are acceptable.

Step 4: Cross-Tabulation

Run Crosstabs for Department × Gender. Results show gender is roughly balanced across departments. The highest proportion of female employees is in HR (71.7%), while IT has the highest proportion of male employees (44.3%).

Step 5: Outlier Check

Boxplots from Explore identify three outliers on work experience (cases 48, 112, 189 — all with 25+ years). Z-scores confirm all three exceed |z| = 3.0. Decision: retain (legitimate values) but note in the report and check whether they influence regression results.

12 Common Mistakes in Descriptive Statistics

Avoid these errors that dissertation committees and journal reviewers flag most often.

1. Using the mean for nominal data

The mean of gender coded 1/2 is meaningless. Use mode and frequencies.

2. Ignoring missing data

Report the N for each variable. If missingness exceeds 5%, discuss its pattern and mechanism.

3. Skipping normality checks

Committees expect normality evidence before parametric tests. Always include Shapiro-Wilk or visual checks.

4. Reporting only "percent" instead of "valid percent"

If you have missing data, the "Percent" column includes missings in the denominator. Use "Valid Percent".

5. Too many decimal places

APA recommends two decimals for means and SDs. Reporting M = 3.4217 is excessive.

6. Pasting raw SPSS output

Reformat SPSS tables into proper APA tables. Raw screenshots are not acceptable in dissertations.

7. No SD alongside the mean

A mean without a standard deviation tells only half the story. Always report both.

8. Treating Likert items as continuous

Individual 5-point items are ordinal. Composite scale scores (summed/averaged) may be treated as interval if reliability is adequate.

9. Ignoring outliers

Report identified outliers, explain how they were handled (retained, Winsorized, removed) and justify the decision.

10. No graphical displays

Tables show numbers; charts show patterns. Include at least histograms and boxplots for key variables.

11. Confusing SD with SE

SD measures variability in your sample. SE (standard error) measures precision of the mean estimate. Descriptive tables use SD.

12. Not linking descriptives to research questions

Don't just dump numbers. Interpret what the descriptives mean in the context of your study's aims.

SPSS Syntax for Descriptive Statistics

For reproducibility, we provide the SPSS syntax commands alongside every analysis. Here are the key commands for descriptive statistics.

Frequencies

FREQUENCIES VARIABLES=Gender Department
  /STATISTICS=MODE
  /BARCHART FREQ
  /ORDER=ANALYSIS.

Descriptives

DESCRIPTIVES VARIABLES=JobSatisfaction WorkExperience Commitment
  /STATISTICS=MEAN STDDEV MIN MAX SKEWNESS KURTOSIS.

Explore (with Normality Tests)

EXAMINE VARIABLES=JobSatisfaction WorkExperience Commitment
  /PLOT BOXPLOT HISTOGRAM NPPLOT
  /COMPARE GROUPS
  /STATISTICS DESCRIPTIVES
  /CINTERVAL 95
  /MISSING LISTWISE
  /NOTOTAL.

Cross-Tabulation with Chi-Square

CROSSTABS
  /TABLES=Department BY Gender
  /FORMAT=AVALUE TABLES
  /STATISTICS=CHISQ PHI
  /CELLS=COUNT ROW COLUMN TOTAL
  /COUNT ROUND CELL.

Z-Score Outlier Detection

DESCRIPTIVES VARIABLES=WorkExperience
  /SAVE
  /STATISTICS=MEAN STDDEV.

* Flag outliers with |z| > 3.29.
COMPUTE outlier_flag = (ABS(ZWorkExperience) > 3.29).
EXECUTE.
FREQUENCIES VARIABLES=outlier_flag.

What You Receive With Every Analysis

1

Complete SPSS Output (.spv)

The full SPSS Output Viewer file containing every table and chart generated during the analysis. Open it in SPSS to explore, pivot or export any table.

2

SPSS Syntax File (.sps)

Every command we ran, documented with comments. You or your supervisor can re-run the entire analysis by opening this file and clicking "Run All" in SPSS.

3

APA Results Write-Up

A Word document with APA-formatted descriptive statistics tables, frequency tables and narrative interpretation — ready to paste into your Chapter 4 or Results section.

4

Charts & Figures

High-resolution (300 DPI) histograms, boxplots, bar charts and Q-Q plots exported as .png files, formatted for APA publication standards.

5

Normality Assessment Report

A summary table of Shapiro-Wilk results, skewness/kurtosis values and a recommendation for each variable: parametric-safe, transform, or use nonparametric alternative.

6

Free Revisions

If your committee requests changes — additional variables, different groupings, alternative charts — we revise at no extra cost until your supervisor signs off.

Frequently Asked Questions

What are descriptive statistics?

Descriptive statistics are numerical summaries and visual displays that describe the main features of a dataset. They include measures of central tendency (mean, median, mode), measures of variability (range, standard deviation, variance, IQR), measures of distribution shape (skewness, kurtosis) and frequency distributions. Unlike inferential statistics, descriptive statistics do not make predictions or test hypotheses — they describe what the data show.

When should I use descriptive statistics in my research?

Every quantitative study should start with descriptive statistics. They appear in Chapter 4 of your dissertation (or the Results section of a journal article) before any inferential tests. Descriptive statistics characterise your sample, verify data quality, check distributional assumptions for later tests and provide context that makes inferential results interpretable. Skipping them is a common committee critique.

What is the difference between descriptive and inferential statistics?

Descriptive statistics summarise and describe your data: mean scores, percentages, standard deviations, charts. Inferential statistics draw conclusions beyond your sample: hypothesis tests (t-tests, ANOVA, regression), confidence intervals and effect sizes. You need descriptive statistics first to understand and present your data, then inferential statistics to test your research questions. Both sections are required in a standard dissertation results chapter.

How do I report descriptive statistics in APA format?

APA 7th edition requires a descriptive statistics table with the variable name, N, mean (M), standard deviation (SD), and optionally the range, skewness and kurtosis. Table format uses no vertical lines and only three horizontal rules (top, header-body separator, bottom). Use italics for statistical symbols: M, SD, N. Report means to two decimal places and standard deviations to two decimal places. Example: 'Participants reported moderate stress levels (M = 3.42, SD = 0.87, N = 215).'

Which descriptive statistic should I use for my variable type?

For nominal variables (e.g. gender, ethnicity): frequencies and percentages. For ordinal variables (e.g. Likert scales, education level): median, IQR, frequencies. For continuous/interval-ratio variables (e.g. age, test scores, income): mean, standard deviation, skewness, kurtosis. If your continuous data are severely skewed, report the median and IQR alongside the mean and SD.

What is normality testing and why does it matter?

Normality testing assesses whether your continuous variables follow an approximately normal (bell-shaped) distribution. It matters because many parametric tests (t-test, ANOVA, Pearson correlation, linear regression) assume normality. SPSS provides the Shapiro-Wilk test (preferred for samples under 2,000) and Kolmogorov-Smirnov test, plus visual checks: histograms with a normal curve overlay, Q-Q plots and P-P plots. If normality is violated, you may need to transform the data or use a nonparametric alternative.

How do I check for outliers in SPSS?

SPSS offers several outlier detection methods: boxplots (cases beyond 1.5× IQR are marked with circles; beyond 3× IQR with asterisks), z-scores (absolute value above 3.29 at p < .001), Mahalanobis distance (for multivariate outliers), Cook's distance (for influential cases in regression). Use the Explore procedure (Analyze → Descriptive Statistics → Explore) to generate boxplots and identify outlier case numbers automatically.

What is cross-tabulation and when do I use it?

Cross-tabulation (crosstabs) displays the joint frequency distribution of two or more categorical variables in a contingency table. Use it when you want to see how categories of one variable relate to categories of another — for example, whether males and females differ in their satisfaction ratings (Low/Medium/High). SPSS cross-tabulation includes row percentages, column percentages, expected counts and chi-square test statistics. It is the descriptive foundation for chi-square tests of independence.

Can you produce publication-ready charts from SPSS?

Yes. We produce APA-compliant histograms, bar charts, pie charts, boxplots, line graphs and scatterplots from SPSS Chart Builder and Legacy Dialogs. Each chart is formatted for publication: proper axis labels, no chartjunk, greyscale-safe colour palettes, appropriate figure titles and notes. We export charts as high-resolution images (.png at 300 DPI) suitable for direct insertion into your Word document or LaTeX file.

How much does a descriptive statistics analysis cost?

Pricing depends on the number of variables, the size of your dataset and the specific analyses required. A basic descriptive profile (frequencies, means, SDs, normality checks) for a standard dissertation dataset is among our most affordable services. Contact us for a free, no-obligation quote — we will review your data and provide an itemised proposal within 24 hours.

Need Help With Your Descriptive Statistics?

Send your dataset and research questions. We'll compute all descriptive statistics, check normality and deliver APA-formatted results within 48 hours.