Start with one file. We’ll show what was read, help you clean and combine waves, and explain the models you choose. You can export the data and report at any point.
1 · What do you want to do?
2 · Add your dataset files
New here? Explore the workflow with a clearly marked simulated dataset.
Use plain English if you already know the constructs. Example: “Combine SUP1–SUP5 into supervisor communication, reverse-code SUP2 and SUP4, then test whether communication moderates the effect of conflict on stress.” You can also choose variables after upload.
We show an audit before you run a model. Some statistical file formats use a server parser; avoid uploading data you are not authorized to process.
Model search without hiding the search.
DataMonster records models tested, unsupported alternatives, effect sizes, confidence intervals, cleaning decisions, and construct scoring. Statistical significance is treated as one piece of evidence—not the objective function.
DataMonster does not label exploratory measurement screens as CFA/SEM, does not claim that Hayes PROCESS itself ran unless a licensed PROCESS environment is connected, and does not silently impute missing values. Those boundaries are intentional so the exported analysis record remains auditable.