Paste your research title or objectives — AI identifies your independent, dependent, moderating, mediating, and control variables with measurement suggestions.
Variables are the building blocks of any quantitative research. Correctly identifying and classifying your variables determines your entire methodology — from research design to statistical analysis. Misidentifying an IV as a DV (or missing a moderator) can invalidate your entire study.
The presumed cause or predictor. What you change/manipulate (experiments) or what you think influences the outcome (correlational). Also called predictor, factor, or treatment variable.
The presumed effect or outcome. What you measure to see if it changed. Also called outcome, response, or criterion variable.
Affects the STRENGTH or DIRECTION of the IV→DV relationship. E.g., "gender moderates the effect of leadership on satisfaction" (the effect is stronger for one gender).
Explains HOW or WHY the IV affects the DV. It's the mechanism. E.g., "motivation mediates the effect of training on performance" (training → motivation → performance).
Moderator = changes the STRENGTH of the relationship (the effect is bigger/smaller depending on the moderator). Mediator = explains the MECHANISM (the IV works THROUGH the mediator). Ask: 'Does it change how strong the effect is?' (moderator) or 'Does the IV work through this variable?' (mediator).
In experimental research, yes — you need to control confounds. In correlational/survey research, control variables are demographic factors (age, gender, experience) that you include in your regression to rule out alternative explanations.
The more specific your title/objectives, the more accurate the identification. Include your population, context, and any suspected relationships. 'Leadership and satisfaction' gives generic results. 'Transformational leadership style on job satisfaction among IT professionals in Bangalore, moderated by organizational culture' gives specific, actionable results.
Yes. Many studies have 2-4 IVs and 1-2 DVs. Multiple regression, MANOVA, and SEM are designed for this. Our tool identifies all variables regardless of count.