Semantic categorization in a linear order problem

Pliske, Rebecca M.; Smith, Kirk H.

doi:10.3758/bf03197603

Cited by 27 publications

(42 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The farther apart two objects are on a dimension, the faster the subject can respond. When items have been learned as explicit linear orders, reaction times (RTs) to pairs of stimuli plotted as a function of the distance of one of the pair members from an endpoint results in a bow-shaped serial position curve (e.g., Maki, 1981;Pliske & Smith, 1979;Woocher, Glass, & Holyoak, 1978). Congruity effects also occur in many studies (e.g., Kosslyn, Murphy, Bemesderfer, & Feinstein, 1977;Maki, 1981).…”

mentioning

confidence: 99%

See 1 more Smart Citation

Why do categorization effects occur in comparative judgment tasks?

Maki

1982

Memory & Cognition

View full text Add to dashboard Cite

Three experiments were conducted to understand what eliminates the symbolic distance effect for reaction times (RTs) when subjects learn that items belong to categories. Artificial city locations were judged, and artificial states served as categories. In Experiment I, whether or not categories were presented and the amount of categorization practice were manipulated. Only those subjects who had practice categorizing cities into states showed a reduced symbolic distance effect for different-state pairs. Analyses of serial position curves indicated that categorization produced faster RTs to different-as compared with their adjacent same-state pairs and it also produced faster RTs for same-state pairs closer to the category border. The border seems to serve as an anchor and facilitates RTs. In Experiments 2a and 2b, distance judgments were made by subjects who either had or had not categorized artificial cities into states. Subjects who learned categories gave larger estimates of distance between cities belonging to different states than to cities belonging to the same state. Categorization seemed to place cities from different states farther apart in psychological space, making their locations more discriminable. This, along with the use of the category border as an anchor, might explain why categorization effects occur in this location judgment task.When people compare two objects on some dimension in memory, several effects are usually observed. One of these has been referred to as a symbolic distance effect (Moyer & Bayer, 1976). The farther apart two objects are on a dimension, the faster the subject can respond. When items have been learned as explicit linear orders, reaction times (RTs) to pairs of stimuli plotted as a function of the distance of one of the pair members from an endpoint results in a bow-shaped serial position curve (e.g., Maki, 1981;Pliske & Smith, 1979;Woocher, Glass, & Holyoak, 1978). Congruity effects also occur in many studies (e.g., Kosslyn, Murphy, Bemesderfer, & Feinstein, 1977;Maki, 1981). Pairs near the end of one dimension are responded to faster with the comparative that corresponds to that end of the dimension than with the other comparative. For example, large items are responded to faster with "larger than" than with "smaller than." In addition to these effects, recent evidence has been found for categorization effects in comparative judgment tasks. If two objects belong to different discrete categories, the symbolic distance effect may not be observed and the serial position curves may show a dip at the category border. The present experiments were designed to investigate such categorization effects.An early demonstration of categorization effects can be found in an experiment by Kosslyn et al. (1977). They had subjects learn names for stick figures of different sizes. Some subjects also learned that the figures Appreciation is expressed to Kim Haroldson for testing some of the subjects in Experiment 1. Requests for reprints should be addressed to Ruth H. Maid, Department of ...

show abstract

mentioning

confidence: 99%

“…Pliske and Smith (1979) had subjects learn a list of names ordered on an intelligence dimension. In one condition, the top half of the list consisted of names of one sex and the bottom half consisted of names of the other sex.…”

mentioning

confidence: 99%

Why do categorization effects occur in comparative judgment tasks?

Maki

1982

Memory & Cognition

View full text Add to dashboard Cite

show abstract

“…Correspondence should be addressed to K. M. Sailor, Department of Psychology, Lehman College, CUNY, 250 Bedford Park Blvd. West, Bronx, NY 10468-1589 (e-mail: kmslc@cunyvm.cuny.edu).tion of categorization effects (Brown & Siegler, 1991;Kosslyn, Murphy, Bemesderfer, & Feinstein, 1977; Maki, 1981;Pliske & Smith, 1979; Sailor & Shoben, 1993). Categorization effects refer to the finding that the symbolic distance effect is either reduced or eliminated if a relevant categorical distinction is correlated with relative magnitude.…”

mentioning

confidence: 99%

“…Categorization effects refer to the finding that the symbolic distance effect is either reduced or eliminated if a relevant categorical distinction is correlated with relative magnitude. For example, Pliske and Smith (1979) had subjects learn a set of names (half male and halffemale) that was ordered on intelligence. In the critical experimental group, gender and intelligence were correlated because all of the names of one gender were brighter than all of the names of the other gender.…”

mentioning

confidence: 99%

The role of categorical information in processing relational attributes

Sailor

Shoben

1996

Mem Cogn

View full text Add to dashboard Cite

In two experiments, the effect of knowledge about relevant common taxonomic categories on judgments of relative magnitude was investigated. In Experiment 1, subjects judged the relative size of objects that belonged to the categories bUilding and animal. The symbolic distance effect was smaller for between-category pairs (e.g., horse-cabin) whose members could be classified more quickly into the two categories. This result strongly suggests that subjects used category membership to determine relative size on at least some portion of the trials. In Experiment 2, a similar attenuation of symbolic distance was obtained when subjects judged the relative height of furniture and vehicles. A simple explanation of the reduced symbolic distance effect for between-category pairs is that subjects occasionally fail to categorize the items into different categories and have to compute relative magnitude from the magnitudes of the two items. Fits to the data using the bootstrap technique suggest that more involved assumptions are required.A familiar issue confronting cognitive psychologists is the problem that the same information can be represented in a variety ofstates that reflect a more or less explicit representation ofsome fact about the world (Smith, 1978). For example, some models of semantic memory assume that category membership is computed from knowledge about the attributes oftheir exemplars, but other models assume that category membership is explicitly represented and retrieved directly (Chang, 1986;Smith, 1978).Within the area of comparative judgment, the predominant view has been that relative magnitude is computed (Banks, 1977;Holyoak, 1978;Moyer, 1973). The main reason that researchers have believed that relative magnitude is computed is the robust phenomenon ofthe symbolic distance effect. The symbolic distance effect is the finding that the time required to select the lesser or greater of two objects increases as they become more similar on a specified dimension. For example, it is easier to determine that a car is larger than a shoe than it is to determine that a car is larger than a piano. This finding suggests that relative magnitude is determined by evaluating differences in the absolute magnitudes ofobjects because there is no reason to believe that response time (RT) would be a function of symbolic distance if the order of a pair was explicitly represented.Subsequently, the notion that relative magnitude is always computed has been challenged with the demonstraThis research was supported, in part, by Grant 2S06GM08225 from NIH to Lehman College and by Grant BN586-08215 from the National Science Foundation to the second author. Correspondence should be addressed to K. M. Sailor, Department of Psychology, Lehman College, CUNY, 250 Bedford Park Blvd. West, Bronx, NY 10468-1589 (e-mail: kmslc@cunyvm.cuny.edu).tion of categorization effects (Brown & Siegler, 1991;Kosslyn, Murphy, Bemesderfer, & Feinstein, 1977; Maki, 1981;Pliske & Smith, 1979; Sailor & Shoben, 1993). Categorization effects refer to the f...

show abstract

“…In recent experiments using artificial (experimenterdefined) linear orders to study inferential reasoning abilities, subjects compared nonadjacent relations faster than adjacent relations (e.g., Pliske & Smith , 1979 ;Potts, 1974;Woocher, Glass, & Holyoak , 1978). Such a "distance effect" has occurred invariably for comparative judgments about linearly ordered information (see Potts , Banks, Kosslyn, Moyer, Riley, & Smith , 1978 , for a review of such effects) .…”

mentioning

confidence: 99%

Encoding partially ordered information

Griggs

Keen

Warner

1980

Bull. Psychon. Soc.

View full text Add to dashboard Cite

The typical distance effect for comparative judgments about ordered information (faster responses for nonadjacent comparisons than for adjacent comparisons) has not been found in recent experiments on processing highly complex 12·term partial orders. However. the rather complicated training and testing procedures used in these studies may have contributed to this failure to observe typical distance effects. The present experiment investigated the processing of a seven-term partial order in simpler experimental conditions similar to those employed in complete order studies. Results indicated that there were pronounced individual differences in encoding such information. Since only three subjects were judged to be using a valid representation of the partial order, no conclusions about the lack of distance effects in comparing partially ordered information could be drawn. Most subjects used memory representations that removed some or all of the indeterminacy in the partial order. It was concluded that most people do not have a p-econceived scheme for encoding partially ordered information. Although highly probable , it is not clear that this indeterminacy in partial order structures is responsible for the failure to observe distance effects in comparative judgment studies using such information. The experimental procedures employed in complete and partial order studies have been very different. The partial order studies have used very elaborate learning and training phases on the adjacent relations before any overall testing. In contrast, the procedures that have been employed in complete order studies are very simple. Subjects usually study either the adjacent relations in some prose frame or the actual complete order for a self-paced period of time. This study period is followed by a true -false or two-alternative forced-ehoice test on all the possible relations in the order .The present experiment investigated the processing of a smaller partial order under experimental conditions like those used in complete order studies. First , the partial order was presented in paragraph form , similar to that used in the initial complete order studies ofIn recent experiments using artificial (experimenterdefined) linear orders to study inferential reasoning abilities, subjects compared nonadjacent relations faster than adjacent relations (e.g., Pliske & Smith , 1979 ;Potts, 1974;Woocher, Glass, & Holyoak , 1978). Such a "distance effect" has occurred invariably for comparative judgments about linearly ordered information (see Potts , Banks, Kosslyn, Moyer, Riley, & Smith , 1978 , for a review of such effects) . However, two studies on the processing of 12-term partial orders have found either no distance effect (Moeser & Tarrant , 1977) or a reverse distance effect-faster responses for adjacents (Hayes-Roth & Hayes-Roth , 1975).Unlike complete linear orders , partial orders cannot be validly arranged in a linear array. Such orders must be represented in a physically multidirectional pattern. A sample partial order structure...

show abstract

Semantic categorization in a linear order problem

Cited by 27 publications

References 11 publications

Why do categorization effects occur in comparative judgment tasks?

Why do categorization effects occur in comparative judgment tasks?

The role of categorical information in processing relational attributes

Encoding partially ordered information

Contact Info

Product

Resources

About