The purposes of this study were to compare the minimum response rates of mailed questionnaires with sincerity answer and nonsincerity answer required for unbiased estimators, and study the impact of nonreponse or nonsincerity answer on estimators' quality. The study covered nine parameters : six arithmatic means and three proportions of background and opinion variables. The population responding to the mailed questionnaires were primary school teachers under the Office of Bangkok Primary Education. The teachers were divided into two groups. The first group comprised the teachers with sincerity answer and nonsincerity answer. The second group is only the teachers with sincerity answer. To check for sincerity, background variables were tested by comparing their answers to a database and opinion variables were tested by Rasch Model. The response rates were set at 50, 55, 60, 65, 70, 75, 80, 85, 90, 95 and 100 percent. The data was carried out using simulation technique with real data. The bias investigation was done by 10 repeated samples at each of the response rates. Interval estimation was employed on each sample using 95% confidence level. The findings were as follows: 1. The response rate and sincerity answer effects on the estimators' quality. 2. The minimum mailed questionnaires reponse rate needed for obtaining the unbias estimators for variables of background was 85 percent for both groups 3. The minimum mailed questionnaires with sincerity answer reponse rate needed for obtaining the unbias estimators for variables of opinion was 70 percent and 80 percent for the group with sincerity and nonsincerity answer. In summary the minimum mailed questionnaires with sincerity answer response rate needed for obtaining the unbias estimators was 85 percent for every variables.
