SPSS command: SPSS META CONTINUOUS
Meta-analysis of continuous outcomes in your browser
A continuous-outcome meta-analysis pools studies that compare the mean of a treatment and a control group.
Free · No account · Files stay on your device · Updated 18 October 2026
What it tells you
Crosstabs runs the SPSS META CONTINUOUS procedure on your device, computing Hedges' g, Cohen's d or raw mean differences from group summaries, or taking precomputed effects and standard errors, then pooling them with inverse-variance fixed or random effects, heterogeneity statistics, Egger's test and a forest plot.
- The pooled standardized or raw mean difference with a confidence interval and p value.
- Each study's effect size, standard error, interval and weight.
- Between-study heterogeneity (Q, τ², H², I² with intervals) under the τ² estimator you choose.
- A test for small-study effects that may signal publication bias.
When to use it
Use it when
- Studies report group sizes, means and SDs of the same outcome, possibly on different scales (use Hedges' g).
- All studies use the same instrument and you want the effect in its natural units (use Mean difference).
- You already have effect sizes and standard errors, e.g. from regression or published tables.
Use something else when
- Outcomes are counts of events: use the binary-outcome meta-analysis.
- You want to relate effect sizes to study characteristics: use meta-regression.
- You have raw data for one study only: use an independent-samples t test.
How to run it in Crosstabs
- 01
Enter one row per study
Prepare a file with one row per study, then open it and switch to Statistics mode.
- 02
Choose the continuous meta-analysis
In the Agreement and diagnostics group, pick Meta-Analysis: Continuous Outcomes (META CONTINUOUS).
- 03
Choose the data form
Set Data to Group means, SDs and sizes (then pick Effect size: Hedges' g, Cohen's d or Mean difference, and map Treatment: N, mean, SD and Control: N, mean, SD) or to Precomputed effect size and standard error (then pick the two columns).
- 04
Pick the model
Set Model to Random effects (choose the Between-study variance (τ²) estimator and optionally Knapp-Hartung adjustment) or Fixed effect, and set the Confidence level (%). Optionally choose a Study label (optional).
- 05
Run
Press Run. Any workspace weight is ignored; studies are weighted by the inverse of their variance.
Worked example
Computed by the Crosstabs engine from the synthetic data below when this page was built — not typed in.
Data
Six hypothetical studies of a reading intervention, each entered as one row with the treatment and control group size, mean and SD of a reading score. These summaries are typed inputs; Hedges' g and every pooled statistic are computed from them.
Input preview
| Study | Treatment: N | Treatment: mean | Treatment: SD | Control: N | Control: mean | Control: SD |
|---|---|---|---|---|---|---|
| Study A | 32 | 54.1 | 9.8 | 30 | 49.6 | 10.4 |
| Study B | 48 | 51.3 | 11.2 | 50 | 50.2 | 10.9 |
| Study C | 25 | 58.7 | 8.9 | 24 | 50.8 | 9.6 |
| Study D | 70 | 52.4 | 10.1 | 68 | 49.9 | 10.3 |
| Study E | 20 | 55.9 | 12.5 | 22 | 48.1 | 11.8 |
| Study F | 41 | 50.9 | 9.5 | 40 | 50.5 | 9.9 |
| Study | Effect Size | Std. Error | 95% CI | Weight % |
|---|---|---|---|---|
| Study A | 0.440 | 0.257 | −0.064 – 0.944 | 14.2 |
| Study B | 0.099 | 0.202 | −0.298 – 0.495 | 20.7 |
| Study C | 0.840 | 0.298 | 0.256 – 1.425 | 11.1 |
| Study D | 0.244 | 0.171 | −0.091 – 0.579 | 26.1 |
| Study E | 0.630 | 0.317 | 0.010 – 1.251 | 10.0 |
| Study F | 0.041 | 0.222 | −0.395 – 0.476 | 17.9 |
Weights are random-effects weights.
| Model | Effect Size | Std. Error | Z / t | Sig. | 95% CI |
|---|---|---|---|---|---|
| Inverse variance | 0.295 | 0.093 | 3.17 | .002 | 0.112 – 0.477 |
| Random effects | 0.310 | 0.107 | 2.89 | .004 | 0.100 – 0.520 |
Random effects: τ² estimated by restricted maximum likelihood (REML).
| Q | df | Sig. | τ² | H² | I² (%) | I² 95% CI |
|---|---|---|---|---|---|---|
| 7.128 | 5 | .211 | 0.0148 | 1.426 | 29.9 | 0.0 – 71.4 |
Q is Cochran's homogeneity statistic for the common-effect estimate; H² = Q/df and I² = (Q − df)/Q.
| Item | Intercept | Std. Error | t | df | Sig. |
|---|---|---|---|---|---|
| Small-study effects | 3.839 | 1.693 | 2.27 | 4 | .086 |
Regression of effect/SE on 1/SE; an intercept different from 0 suggests funnel-plot asymmetry.
How to read it
The random-effects pooled Hedges' g is 0.310 (95% CI 0.100 – 0.520, p = .004), a small-to-moderate benefit of the intervention. Between-study variance is τ² = 0.0148 with I² = 29.9%, so the studies are about as consistent as sampling error allows. Egger's intercept (3.84, p = .086) is read with caution: with six studies the test has little power.
Output tables explained
- Effect Size Estimates
- Per study: effect size, standard error, confidence interval and percentage weight. Footnotes give the variance formula used and list studies excluded for fewer than two cases per group or zero SD.
- Overall Effect Size
- Inverse-variance fixed-effect and, for random effects, the random-effects pooled estimate with SE, Z or t, significance and interval.
- Heterogeneity
- Q, df, significance, τ², H² and I² with intervals.
- Egger's Regression-Based Test
- The intercept of the regression of effect/SE on 1/SE with its t test on k − 2 df; a significant intercept suggests asymmetry.
Assumptions and notes
- Within each study the outcome is roughly normal so means and SDs summarise it well.
- Standardized effects assume the studies measure the same construct, even if on different scales.
- Hedges' g and Cohen's d standardize by the pooled SD; Hedges' g applies the exact small-sample correction.
- Precomputed rows need a positive standard error; rows with zero or missing SE are dropped.
How the results are validated
On a seven-study reference set, Hedges' g and mean-difference effect sizes and variances match R metafor 5.2.1 escalc to 1e-12, and all seven τ² estimators with their pooled estimates match rma to 1e-6 for interior solutions; Hedges' g also agrees with statsmodels 0.15 effectsize_smd, and the DerSimonian-Laird τ² with statsmodels combine_effects, to 1e-3. Egger's test matches metafor regtest to 1e-8. Validation evidence
Questions
Hedges' g or Cohen's d?
Hedges' g corrects the small upward bias of d in small studies and is the usual choice in meta-analysis. With groups of 30 or more the two are almost identical.
Can I combine change scores and post-test scores?
Only with care: their SDs differ, so standardized effects are not directly comparable. Prefer one type, or compute effects and SEs consistently yourself and use the precomputed option.
How many studies do I need for Egger's test?
The panel shows it from three studies, but the test has little power below about ten. Treat a non-significant result with few studies as uninformative, not as evidence of no bias.
What does the Knapp-Hartung adjustment do?
It rescales the random-effects standard error and uses a t distribution with k − 1 df, giving wider, better-calibrated intervals when there are few studies.
Related procedures
- Meta-Analysis: Binary OutcomesMETA BINARY
- Meta-RegressionMETA REGRESSION
- Independent-Samples T TestT-TEST GROUPS
- One-Way ANOVAONEWAY
- Free SPSS alternative: full coverage list
Run Meta-Analysis: Continuous Outcomes on your own data
Open a .sav, CSV or Excel file. Everything runs in your browser; the file never leaves your device.
SPSS is a trademark of IBM Corp. Crosstabs is not affiliated with or endorsed by IBM.