JOURNAL OF EXI146fRIMFNIAI CHII I I146S110101 31 370186 I - PDF document

1 JOURNAL OF EXI’f~RIMFN~IAI CHII I)
JOURNAL OF EXI’f~RIMFN~IAI CHII I) I’S\11010(~1 31. 370-186 (I%!) Developmental Trends in the Use of Perceptual and Conceptual Attributes in Grouping, Clustering, and Retrieval RACHEL. MELXMAN BARBARA TVERSK’L ,\ N �I DAPHNA BAKA I‘/ In Experiment I. been conflicting evidence on the ability of young children to utilize superordinate taxonomic categories. such as vehicle or This research was suppor

2 ted by grants t’t-om the Human Deve
ted by grants t’t-om the Human Development Institute at the Hebrew University of Jerusalem and from the Faculty of Humanities and Social Sciences ot Ben Gurion University and National Instituteof Mental Health Grant MH-3424X-01 toStanford University. Preparationofthe manuscript PERCEPTUAL VS CONCEPTUAL ORGANI%ATION 471 to organize memory (Lange, 1978; Moely, 1977). On the one hand, there is little, if any, unambiguous evidenc

3 e that children under 6 spontaneously us
e that children under 6 spontaneously use superordinate category for clustering (Lange, 1978). On the other hand, experimentally imposed cues for young children (Eysenck & Baron, 1974). One possible resolution of these findings comes from the observation that the experiments showing use tasks emphasizing errc.ocii!rg aspects of memory, while those experiments finding a beneht of superordinate categories use tasks emphasizing rrtr

4 kwl aspects of memory. So, whereas youn
kwl aspects of memory. So, whereas young children may not spontaneously encode stimuli together by category, they may nevertheless. use superordinate categories as retrieval cues. There is considerable evidence occurs predominately on the basis of perceptual features for young children and conceptual features for older successes and failures in recall. According to the encoding specificity hypothesis. recall is facilitated cues

5 may augment effects of initial encoding
may augment effects of initial encoding (Moscovitch & Craik, 1976: Tulving & Watkins, 1975). and that conceptual retrieval cues may be superior to perceptual ones even when initial encoding was not conceptual (Moscovitch Sr Craik. 1976). 472 MELKMAN.TVERSKY.ANDBARATZ The preceding analysis suggests that young children may spontaneously organize, but may prefer perceptual however, because of the effectiveness in retrieval of concep

6 tual, semantic attributes as opposed to
tual, semantic attributes as opposed to superficial, perceptual ones, imposed conceptual organization may facili- tate even young children’s performance. This analysis PERCEPTUAL VS CONCEPTUAL ORGANIZATION 473 EXPERIMENT I Method Subjects. Thirty Israeli middle- and upper- middle-class children from each of three age groups, preschool, kindergarten, and fourth grade. served as subjects. These ages were selected on the basis

7 of previous research to cover as much of
of previous research to cover as much of Stimuli. Amrrican Heritage Word Frequrncy Book (Carroll, Da- vies. and Richman, 1971) count for children’s literature were 52 for the color list, 56 for form, and for concept. The grouping task consisted of 63 colored drawings, again of common objects, divided into three sets of triads, in order to assess separately 474 MELKMAN. TVERSKY. AND BARATZ preferences for color versus form.

8 color versus concept, and form versus co
color versus concept, and form versus concept. As before, colors and forms were selected to be natural and typical of the objects. One drawing of each triad served Color match Anchol Blackboard (Top) hat Umbrella Tire F’orm match Cake Drum Color match Tree Fire engine Anchor Frog Kose Category match Monkey Sunflower Form match Bow Pencil Anchor ButterRy Flute Category match Fly Accordion PERCEPTUAL VS CONCEPTUAL ORGANIZATION 4

9 75 set was presented to a Rcsrrlts Gro
75 set was presented to a Rcsrrlts Grouping task. For each child, the number of color, form. and concept choices was determined for each set of the grouping triads (color versus form. color versus concept, and form versus concept) and for each pre- sentation separately. The MELKMAN. TVERSKY. AND tiARAT% of form choices, significant effects of comparison type (k“( 1, 87) = 13.X. 17 .Ol) and of the age by type interaction (F(?

10 , closest to it in the hypothesized or
, closest to it in the hypothesized ordering (i.e., form) gained in prefere rice. A more stringent test of the developmental progression of the three features as bases for grouping was provided by examination of the prefer- ence patterns of individual PERCEPTUAL VS CONCEPTUAL ORGANIZATION 477 The preference protocol of each child was assigned to one of the six dominance categories or to the intransitive category, or to the no domi

11 - nance category where no strict orderin
- nance category where no strict ordering was evident (e.g.. where any pair TABLE 3 FREQUENCIES OF PREFERENCE PATTERNS IN GROUPING 47x MELKMAN. TVERSKY. AND BARAT% List Color Free recall Cued recall Cued recall as a percentage ot unrecalled items Total recall Form Free recall Cued recall Cued recall as a percentage of unrecalled item3 Total recall Concept Free recall Cued poorer on the form list than on the other two lists at all

12 ages. For cued recall. only the effect o
ages. For cued recall. only the effect of cue type was significant PERCEPTUALVSCONCEPTUALORGANIZATIC)N RA.~IO OF REP~IITION (RR) SCORES BY LISI TYPE AK:D AGF IN EXPERIMENT I each mean is based, for the different groups. A 3(age) by 3(list) analysis of variance with list type as a within-subject factor yielded no significant effects for age or list but a significant age by list interaction (F(4. 480 MELKMAN. TVERSKY. .4ND BARATZ

13 TABl,E 6 MEN NUMHER ot‘CArEc;oKrEs
TABl,E 6 MEN NUMHER ot‘CArEc;oKrEs AND Wows PLR CAIEGORY R~.C,NLCD. BY AAIU’D IN ENPt:RIMFI\Il 1 List Categories Color Form shift in the bases for clustering in free recall. Younger children grouped objects sharing color or form in preference to grouping objects sharing taxonomic category, but oldest children showed the opposite prefer- ence. Similarly, Sritmrli. Three new sets of drawings were prepared. as follows: PER

14 CEPTUAL VS CONCEPTUAL ORGANIZATION 481
CEPTUAL VS CONCEPTUAL ORGANIZATION 481 Color list: Banana. key, butterfly, 481 MELKMAN. TVERSKY. AND BARATZ Lisl COIOI Free recall Cued recall Cued recall as a both individual patterns et al., 1963: Melk- man bi Deutsch, 1977: Melkman et al.. 1976: Olver & Hornsby. 1966: Suchman Kr Trabasso, 1966) by demonstrating transitivity of preferences PERCEPTUAL VS CONCEPTUAL ORGANIZATION 483 within children as well as evidence for the

15 developmental shift with famil- iar obje
developmental shift with famil- iar objects in their natural shapes and colors. The findings for the recall task are more complicated, and seem to result from the interplay of two factors, one developmental factor affecting 484 MELKMAN. TVERSKY. AND BARATZ orientation of the viewer. must be abstracted over different points of view and over time. The superordinate category of an object is not given in perception at all. and so requi

16 res relatively more abstractiom than att
res relatively more abstractiom than attri- butes given in perception. As for predictability, the Bousheld, W. A. The occurrence of clustering in the recall of randomly arranged associates. Journul ot Gunrrul P.Y,dlo/og~. 19.53. x9-‘40. Brian, C. R.. & Goodenough. F. L. The relative potency of color and form perception at PERCEPTUAL VS CONCEPTUAL ORGANIZATION 485 Cole, M.. Frankel, F., & Sharp, D. Development of free recall l

17 earning in children. De- \~elopmrntul Ps
earning in children. De- \~elopmrntul Psychology. 1971, 4, 109-123. Craik, F. 1. M.. & Jacoby. L. Elaboration and distinctiveness in episodic memory. In G. Nillson (Ed.), Perspectives on mrn)o~ 486 MELKMAN. TVERSKY. AND BARAT% Perlmutter. M.. Cyr Ricks. M. Recall in pre-school children. Jorrrrr~rl c!/ E’~lwriw~e/~ttr/ ~‘/ril,l Psyho/o~~~, 1979. 27, 4’3-436. Posner. M. I. Abstraction and the proces, HLVISEI): July 23