Though you concede that the results of your survey "may not be 100 per cent reliable" (Special Edition, Mar. 1-7), you seem satisfied that the results are reliable enough to warrant publication.

I’m not comfortable with the implications that the percentage figures reported are more or less satisfactory estimates for the percentage figures for the larger population -- "readers of the Mr. & Ms. special edition and those interested in the survey" - of which your sample is presumably representative.

Your report (Jan. 25-31) that 600 of the respondents were students or employees of the De La Salle University whose votes were solicited by Ms. Nilda Rotor. Constituting 9 per cent of your respondents, this group of DLSU-people-to-whom-Ms.-Rotor-has-direct-access is grossly over represented in your sample. On this basis alone, your sample can’t be representative of "readers of the Mr. & Ms. special edition and those interested in the study."

Every decent survey report includes some statement to the effect that we can be 95 per cent sure that the true percentage figures for the population are equal to the corresponding percentage figures in the sample, plus or minus so and so many percentage points.

The statistical analysts you hired (Jan. 25-31) should furnish this margin of error.

Ruben Muego