This paper discusses a longitudinal, exploratory study of developmental dimensions related to object permanence theory and explains how multidimensional scaling techniques can be used to identify developmental dimensions. Eighty infants, randomly assigned to one of four experimental groups and one of four counterbalanced orders of stimuli, were tested at 4 and at 8 months of age. Each group was tested on six cloth hiding tasks spanning several stages (18 trials) and seven with-occlusion and without-occlusion tracking tasks. The four experimental groups, differing in how much information about the object was available, were Regular, Sound, Intermittent Sound, and Control. Infants were ordinally assessed from videotapes with adapted measures designed to make haptic and visual assessment parallel. One way analysis of variance revealed significant differences in performance both among and between experimental groups. Guttman scales were obtained for all groups except the control group. Use of non-metric techniques revealed that the task items did not scale as predicted in traditional object permanence theory. Additional statistical techniques were used to test a Piagetian object permanence dimension interpretation. These techniques showed that individual differences were not important. A Piagetian interpretation was not supported. It is concluded that the multidimensional scaling approach is useful, with qualifications, and can be used to supplement traditional pass/fail Guttman scales. (Author/RH)