DrAIStudio 1
Upload your quantitative data
CSV · XLSX · XLS · TXT · DAT
Drop your file here or
·
Rows = respondents, columns = variables. Name scale items with a prefix + number (e.g. TL1, TL2 … JS1, JS2) for automatic construct detection.
2
State your hypotheses
plain English — one per line📖 How to phrase each hypothesis type (guide)
| Analysis you want | How to write the hypothesis | What SmartSEM runs |
|---|---|---|
| Direct effect | "X positively/negatively affects Y." e.g. Transformational leadership positively affects job satisfaction. | Structural path X → Y with β, p, CI, f² |
| Simple mediation | "M mediates the relationship between X and Y." | Paths X→M→Y (+ direct X→Y), bootstrapped a×b, VAF, Zhao typology |
| Serial mediation | "M1 and then M2 serially mediate the relationship between X and Y." or: "X affects Y through M1 and then M2." | Hayes model-6 path set; three specific indirect effects incl. X→M1→M2→Y |
| Parallel mediation | "M1 and M2 mediate the relationship between X and Y." (no "then" — both mediators work side by side) | Specific indirect via each mediator + contrast test (a₁b₁ − a₂b₂) |
| Moderation | "W moderates the relationship between X and Y." | X×W interaction, simple slopes ±1SD, Johnson–Neyman, 3D surface |
| Moderated mediation (conditional indirect effect) | "W moderates the indirect effect of X on Y through M." Add "(second stage)" if W moderates M→Y instead of X→M. | Index of moderated mediation + conditional indirect effects at W = −1SD / mean / +1SD (Hayes models 7 / 14) |
Tip: one hypothesis per line, starting with H1:, H2:, … Name constructs consistently — SmartSEM maps them to your item prefixes (e.g. "Job Satisfaction" → JS1…JS4) automatically.
Detected hypotheses (edit anything that was misread)
3
Review the model
constructs ↔ your columnsConstruct measurement mapping
| Construct | Mapped items | k | α |
|---|
Conceptual model
4
Run the full analysis
Endogeneity check (instrumental variable, optional)
2SLS with Durbin–Wu–Hausman endogeneity test, weak-instrument F and Sargan test (2+ instruments). Instruments must affect Y only through X. Control variables are used as covariates automatically.
Everything runs locally — large bootstraps may take a minute.
Results
Extra bivariate tests (on demand)
Ctrl/Cmd-click to select several variables (items, composites or numeric columns).