THE HEAD START EVALUATION AT ISSUE EMPLOYED A CAUSAL MODELING TECHNIQUE AND FOUND POSITIVE EFFECTS FOR A SAMPLE OF HEAD START CHILDREN FOR WHOM NEGATIVE EFFECTS HAD BEEN FOUND BY OTHER INVESTIGATORS. A CRITIQUE OF THE EVALUATION, NOTING THAT NONE OF THE ESTIMATES OF EFFECTS (ALL OF WHICH WERE POSITIVE) WAS SIGNIFICANTLY DIFFERENT FROM ZERO (NO EFFECT) AT THE 0.05 LEVEL OF SIGNIFICANCE, CONCLUDES THAT HEAD START EFFECTS ARE NOT YET DEMONSTRABLE. IT IS SUGGESTED THAT THE CRITIQUE OVEREMPHASIZES THE IMPORTANCE OF THE STATISTICAL TEST, PARTICULARLY THE RIGID ADHERENCE TO THE 0.05 LEVEL OF SIGNIFICANCE. OTHER PROBLEMS ARE SAID TO RELATE TO OVERSTATED AND MISLEADING ASSERTIONS IN THE CRITIQUE. RESPONSES TO SPECIFIC ASSERTIONS ARE PRESENTED, AND THE PURPOSE OF THE HEAD START EVALUATION IN QUESTION IS CLARIFIED. THE MODEL EMPLOYED IN THE EVALUATION AND REEXAMINED IN THE CRITIQUE IS DISCUSSED. TO ENCOURAGE FURTHER ANALYSIS, ADDITIONAL DATA ON THE SUMMER HEAD START PROGRAMS (MEANS, STANDARD DEVIATIONS, AND CORRELATIONS FOR HEAD START AND COMPARISON CHILDREN, DATA ON BLACK CHILDREN) ARE PRESENTED. A LIST OF REFERENCES IS INCLUDED. SEE NCJ-50186 FOR THE CRITIQUE. (LKM)