Among several approaches to evaluating measurement quality, the true score MTMM model proposed by Saris and Andrews (1991) was selected, owing to its ability to estimate reliability, validity and method effect of each variable seperately. On the basis of theoretical and empirical findings, four netowork generators were designed to measure four theoretically firmly established social support dimensions. The contest of network generators measuring social companionship and material, informational, and emotional support exchange were adjusted to the characteristics of the experimental groups - eight third-year classes in Gimnazija Bežigrad in Ljubljana. A complex experimental design to study the effects of measurement instruments on the measurement quality was proposed. All the variables included in the experimental design were previously found to affect measurement quality either by survey methodologists or by social network methodologists. Experimental variations included the provision of four response scales (a binary scale, a five point ordinal scale with labeled extreme values, a five point ordinal scale with labeled all values and a line drawing scale), two data collection techniques (recognition and a free recall), two directions of support exchange (original and reversed questions) and varying elapsed times between repeated measurements (20 minutes or one week).