Hum. Dev. 21: 334-345 (1978)
Perceptual Sensitivity and Conceptual Coordination in Children and Younger and Older Adults1 Robin L. West, Richard D. Odom and Jean R. Aschkenasy Department of Psychology, Vanderbilt University, Nashville, Tenn.
Key Words. Aged • Cognitive development • Developmental age groups • Perceptual development • Perceptual discrimination • Problem solving • Visual perception Abstract. This study was designed to explore the effects of perceptual salience on performance in problems requiring the coordination of information. Groups of children, younger adults, and older adults were administered a salience-assessment task to determine each individual’s perceptual sensitivity to each of three dimensions. Half of the subjects in each age group were given a coordination problem with their two most salient dimensions relevant. The remaining half were given problems with their two least salient dimensions relevant. For each of the age groups, those problems containing the most salient information were solved faster and more accurately than problems containing the least salient informa tion. The results demonstrate that perceptual sensitivity continues to influence problem solving from childhood through late adulthood.
The ability to combine or coordinate information from more than one source is essential for solving a number of different problem-solving tasks. For example, in matrix problems, the information presented in both the columns and the rows must be coordinated in order to choose the appropriate stimulus compound to Fill an empty cell in the matrix (e.g., Overton and Jordan, 1971). Also, in recall problems where the subject is required to remember the location of one of several stimulus compounds in an array, the subject must integrate
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
1 This study is based on a thesis submitted by the first author to the Graduate School of Vanderbilt University in partial fulfillment of the requirements for the master’s degree. The authors are indebted to the staff of the Memphis State University Campus School and the Memphis Jewish Community Center whose cooperation made this study possible.
335
values representing different dimensions or relations in the array (e.g., Cun ningham and Odom, 1978). Similarly, conjunctive classification problems are solved by learning to choose the stimuli containing the appropriate values from the relevant dimensions in the problem (e.g., King, 1966). Previous research has shown that the perceptual salience of the relevant information influences performance on coordination problems. To account for this finding, it has been proposed that the perceptual system’s sensitivity to information in such problems plays an important role in determining how rapidly conceptual coordination will occur (e.g., Odom et al, 1975). A distinc tion is made here between (a) perceptual processes associated with the detection of and sensitivity to information in a problem, and (b) conceptual processes associated with evaluation of the information’s relevance for problem solution, e.g., processes like coordination, analysis, categorization, etc. From this view point, information is characterized as relations or dimensions of difference or invariance that are directly perceived without the mediation of conceptual processes (Gibson, 1969). Gibson has proposed that an increasing number of relations are discovered or detected with increasing age. Odom and Guzman (1972) have further pro posed that even after relations are detected, the perceptual system is dif ferentially sensitive to them, that is, some relations are more perceptually salient than others. They have posited that the more salient information in problem solving tasks should have a higher probability of being conceptually evaluated than the less salient information, regardless of which is relevant for problem solution. Support for this view is provided by the finding that coordination problems are solved more rapidly when the relevant information is high in salience than when it is low in salience (Astor-Stetson e ta l , 1978; Cunningham and Odom, 1978; Katsuyama and Hoffarth, 1978; Odom et al, 1975; Overton and Jordan, 1971). Since older children generally solve problems more accurately and rapidly than younger children, many developmental psychologists have argued that older children have certain conceptual abilities which younger children do not have (e.g., Inhelder and Piaget, 1964). Odom and his colleagues entertain the alter native possibility that observed age differences in problem solving may be due largely to changes in perceptual sensitivity. They propose that older children, having had more experience with a variety of relations in the environment, are more sensitive to more relations than younger children. Consequently, in a given problem, the older child should be more likely than the younger child to evaluate more relations and thus should be more likely to solve the problem,
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
Perceptual Sensitivity and Conceptual Coordination
West/Odom/Aschkenasy
336
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
thereby demonstrating the presence of a particular conceptual ability. If this is the case, then age-related differences in problem solving would be due to differences in perceptual sensitivity rather than differences in conceptual pro cessing. Support for this view is provided by the evidence that younger children can solve problems that require the use of processes which conceptual-change theorists have attributed only to older children (Odom, 1978). For example, younger children solving coordination problems containing relevant information that is high in salience perform as well as older children solving conceptually identical problems containing relevant information that is low in salience (Cun ningham and Odom, 1978; Katsuyama and Hoffarth, 1978; Odom et al., 1975). There is considerable evidence from salience research with children as old as 12 years of age that the perceptual system continues to play an important role in problem solving throughout childhood. Nevertheless, there is a popular view expressed in the literature that the role of the perceptual system decreases with development while the role of the conceptual system increases (e.g., Wright and Vlietstra, 1975). That is, the younger child’s problem-solving performance is thought to be controlled by perceptual factors, and the older child’s perfor mance is thought to be controlled by the conceptual or logical requirements of a task. In contrast, the perceptual-salience position stresses the importance of the perceptual system throughout the life span. Given that the information in a problem must be perceived before it can be conceptually evaluated, such an emphasis would be tenable even in the absence of empirical support. It follows that at any point in the life span, the perceptual system’s sensitivity to informa tion should affect the probability that the information will be conceptually evaluated. To explore the influence of perceptual sensitivity on the development of coordination skills across the life span, a salience-assessment task and a coordina tion task were administered to younger and older adults as well as to children. The coordination task was quite different from those used in previous studies. Each subject was shown a series of items, each consisting o f three cards containing stimulus values representing two dimensions. The task required co ordination because the correct card (e.g., a red square) contained one dimen sional value that matched a second card (e.g., a red circle) and another dimen sional value that matched a value on a third card (e.g., a blue square). The task was presented in three phases which were designed to reduce the task require ments and increase the salience o f the relevant information over time. In phase I, single items were presented for the subject’s response. In phase II, the subject
Perceptual Sensitivity and Conceptual Coordination
337
was shown three items with the correct responses indicated, and was asked to respond to a fourth item. It was hoped that the change from a successive (phase I) to a simultaneous (phase II) presentation would reduce the memory requirements of the task. By looking at several solved items presented simul taneously, it was possible for the subject to eliminate certain inappropriate hypotheses before making a response to the unsolved item. In phase III, those subjects who did not solve the task in an earlier phase were told that the task required the coordination of certain unspecified information.
Method Subjects The participants in this study were 40 children, 10-14 years of age (mean = 11.9 years, SD = 0.74); 40 younger adults, 18-22 years of age (mean = 19.7 years, SD = 1.56); and 40 older adults, 60-78 years of age (mean = 69.4 years, SD = 4.93). Half of the individuals at each age level were male and half female. The children, 16 blacks and 24 whites, were enrolled in a predominantly middle-class school in Memphis, Tenn. The younger adults were from psychology classes at Vanderbilt University. The older adults were community-dwelling volunteers from senior centers and clubs in Memphis, Tenn. Three fourths of the older adults had completed a high school education, and their mean number of completed years of formal education was 13.1. All adult participants were white.
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
Materials Salience Assessment. The assessment task was used to determine the relative salience of the dimensions of form, color, and position. The task v/as a modified method of triads similar to the one used by Odom and Guzman (1972). Each salience-assessment item consisted of a 27.2 X 35 cm black card on which three 11.5 X 11.5 cm white cards were placed. Each white card had pictured on it a compound of four values, one from each of four stimulus dimensions. Color was represented by red, blue, green, yellow, orange, and purple. The forms were cross, square, circle, inverted T, triangle, and pentagon, each one no larger than 1.3 cm in diameter. Position values were represented by placement of stimuli either across the upper right to lower left diagonal, the upper horizontal edge, the right vertical edge, the vertical midline, the upper-left-to-lower-right diagonal, and the horizontal midlinc. The number values were 2, 3, 4, 5 ,6 , and 7. On all items, the number values were different on all three cards. There were four practice items. On the first practice item, the values of the three dimensions of form, color, and position were identical on two cards and differed on the third card. On each of the remaining practice items, one pair of cards had identical values on two dimensions while the values of the third dimension were different. The remaining card on these items differed from the other two cards on all dimensions. Each of the salience-assessment items contained three cards, two of which had identical values from one dimension and two of which had identical values from another dimension.
West/Odom/Aschkenasy
338
All three cards contained different values on the third dimension. Thus, two pairs of cards on each item had identical dimensional values. The three possible two-dimensional compari sons (form-color, form-position, color-position) were each presented six times. Problem-Solving Task. The problem-solving task was constructed from the same dimen sional values as those used during the salience-assessment procedure except that the value for number was always 3 and, therefore, number was held constant for all subjects. On each item of the task the subject was shown three cards side by side. Each of the 18 items was composed of a unique group of three cards. On each card was a compound stimulus com prised of values from several dimensions. There were three problems: (a) a form-color problem with form and color variable and relevant (position was held constant and was represented by placement of the stimuli on the horizontal midline); (b) a position-color problem where all the forms were squares, and (c) a form-position problem with all red stimuli as shown in figure 1. Each of the six values from the two relevant dimensions occurred three times in the 18 items. The correct card was randomly placed across trials with the restriction that it occurred an equal number of times in the left, right, and middle positions.
Fig. 1. Example of phase l (item d alone) and phase II (items a-d , with the correct card marked a on items a-c) of the coordination problem.
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
Procedure The experimenter began the salience-assessment task by showing the subject the first practice item, pointing to the two cards that were most similar and saying: They are more like each other than this card (pointing to the third card) which is different from these (pointing to the two most similar cards). On every page in this stack
Perceptual Sensitivity and Conceptual Coordination
339
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
there will always be cards that are more alike. You look at the cards on each page and point to the cards that are more alike. The experimenter then administered the next three practice items followed, without interruption, by the 18 salience-assessment items. The subject’s choice for each item was recorded. The frequency of choices based on each dimension determined the subject’s salience hierarchy, in which the most salient dimension was A (the most frequently chosen) and the least salient dimension was C. If a subject picked two dimensions an equal number of times, those two dimensions were designated an AB tie or a BC tie. After the experimenter determined the individual’s salience hierarchy, the problem solving task was administered. 20 subjects at each age level were assigned to two groups one group was given a task where the subjects’ A and B dimensions were variable and relevant and the C dimension was constant (the AB condition) while the other group received a task where the subjects' B andC dimensions were variable and relevant while the A dimension was held constant (the BC condition). If a subject had an AB tie on the salience-assessment task, he was assigned to the AB condition and if he had a BC tie he was assigned to the BC condition. All other subjects were randomly assigned to the two condi tions. Since the placement of the cards in the array (left, right, or middle) was variable irrelevant information, subjects were told in the instructions that placement was irrelevant. Thus, there were only two variable relevant dimensions in each problem. As the experimenter gave the task instructions, she showed the subject a sample item. The three cards in the sample items contained human cartoon characters taken from maga zines. These characters varied in color, size, shape, apparent age, type of clothing, direction the character faced, etc. The subjects were given the following directions: Today I am going to give you a learning task. I am going to show you a large number of pictures, three at a time, arranged like this (pointing to the sample item). Each time I show you three pictures, one of the three pictures will be the correct picture, according to a rule. You will need to discover the rule. You will have to guess at first, but later you will discover the rule and you will be able to pick the correct picture all the time. The experimenter then gave an example of a rule for the sample item, and continued with the instructions: The rule has nothing to do with where the correct picture occurs in the set. Now each time I show you three pictures, you point to the picture which you think is the correct picture. Then I will tell you which picture is correct. Pay close attention to all the pictures so you can figure out the rule which tells you which pictures are correct. Once you know the rule, you will always be able to pick the correct picture in each set of three pictures. The solution rule for the concept identification problem was this: pick the card which has the same value on one dimension (B) as another card in the set of three and the same value on the second dimension (A or C) as the remaining card. For example, in the formposition problem in figure 1, item d, the correct card (3) matched the form value in card 2 and matched the position value in card 1. In phase I of the task, each of the 18 items was presented one at a time for the subject’s response. After each response, the experimenter indicated the correct card by pointing to it and saying ‘Yes (No), this is the correct picture’. If the subject did not reach the solution criterion (six consecutive correct responses and an explanation of the rule) during the initial presentation of the 18 items, the second phase of the task began.
West/Odom/Aschkenasy
340
In phase II the experimenter presented the original 18 items four at a time. The subject was told: Now 1 am going to show you four sets of pictures at the same time. I will give you the answer for the first three, then you try and pick the correct picture for the fourth set. The rule has to do with each set of three pictures by itself I’m only doing this so you can see several examples at one time. Remember, the rule has nothing to do with whether the correct picture is on the left, right, or in the middle. The first four items were placed before the subject, and the experimenter put a token on the correct cards for the first three items, saying ‘this is the correct picture here’ (see fig. 1 for an example of the form-position problem). The subject was then asked to pick the correct card for the fourth item. After the response and feedback, the experimenter picked up items 1 -4 and similarly placed the next four items in front of the subject. This proce dure continued for six sets of four items, or until the subject reached a criterion of three consecutive correct responses and an explanation of the correct rule. If the subject did not solve the problem with this procedure, phase III began. In phase III the experimenter told the subject, ‘The rule has to do with how the pictures are alike. So look closely to see how the pictures are alike.’ Three more sets of four items were presented as in phase II, and then the experimenter provided another hint: ‘There are two ways that the pictures are like each other. The rule has to do with two ways that the pictures are alike.’ If the subject did not reach criterion after three more sets of four items were presented, the experimenter provided more hints: ‘How is this picture (pointing to the last correct card) like the other two pictures here?' This final procedure was necessary in four cases for the sixth graders and four cases for the older adults. One sixth grader and one older adult in the BC condition did not solve the problem after six more sets of four items were presented and these two subjects were eliminated from the data analysis. There were three dependent measures recorded for each subject: (a) the number of trials to criterion; (b) the percentage of errors, i.e., the number of errors divided by the number of trials to criterion, and (c) the type of errors, whether errors based on form, color, or position. For example, a form error occurred when the subject in the form problem picked the incorrect card which matched the correct card on form.
Results
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
The mean number of trials to criterion and the mean percentage of errors for each age group in the AB and BC problems are presented in table I. A 2 (condition) X 3 (age) analysis of variance was performed on the dependent variable of trials to criterion. There was a main effect for condition, F (l, 114) = 16.8, p
Perceptual Sensitivity and Conceptual Coordination in Children and Younger and Older Adults1 Robin L. West, Richard D. Odom and Jean R. Aschkenasy Department of Psychology, Vanderbilt University, Nashville, Tenn.
Key Words. Aged • Cognitive development • Developmental age groups • Perceptual development • Perceptual discrimination • Problem solving • Visual perception Abstract. This study was designed to explore the effects of perceptual salience on performance in problems requiring the coordination of information. Groups of children, younger adults, and older adults were administered a salience-assessment task to determine each individual’s perceptual sensitivity to each of three dimensions. Half of the subjects in each age group were given a coordination problem with their two most salient dimensions relevant. The remaining half were given problems with their two least salient dimensions relevant. For each of the age groups, those problems containing the most salient information were solved faster and more accurately than problems containing the least salient informa tion. The results demonstrate that perceptual sensitivity continues to influence problem solving from childhood through late adulthood.
The ability to combine or coordinate information from more than one source is essential for solving a number of different problem-solving tasks. For example, in matrix problems, the information presented in both the columns and the rows must be coordinated in order to choose the appropriate stimulus compound to Fill an empty cell in the matrix (e.g., Overton and Jordan, 1971). Also, in recall problems where the subject is required to remember the location of one of several stimulus compounds in an array, the subject must integrate
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
1 This study is based on a thesis submitted by the first author to the Graduate School of Vanderbilt University in partial fulfillment of the requirements for the master’s degree. The authors are indebted to the staff of the Memphis State University Campus School and the Memphis Jewish Community Center whose cooperation made this study possible.
335
values representing different dimensions or relations in the array (e.g., Cun ningham and Odom, 1978). Similarly, conjunctive classification problems are solved by learning to choose the stimuli containing the appropriate values from the relevant dimensions in the problem (e.g., King, 1966). Previous research has shown that the perceptual salience of the relevant information influences performance on coordination problems. To account for this finding, it has been proposed that the perceptual system’s sensitivity to information in such problems plays an important role in determining how rapidly conceptual coordination will occur (e.g., Odom et al, 1975). A distinc tion is made here between (a) perceptual processes associated with the detection of and sensitivity to information in a problem, and (b) conceptual processes associated with evaluation of the information’s relevance for problem solution, e.g., processes like coordination, analysis, categorization, etc. From this view point, information is characterized as relations or dimensions of difference or invariance that are directly perceived without the mediation of conceptual processes (Gibson, 1969). Gibson has proposed that an increasing number of relations are discovered or detected with increasing age. Odom and Guzman (1972) have further pro posed that even after relations are detected, the perceptual system is dif ferentially sensitive to them, that is, some relations are more perceptually salient than others. They have posited that the more salient information in problem solving tasks should have a higher probability of being conceptually evaluated than the less salient information, regardless of which is relevant for problem solution. Support for this view is provided by the finding that coordination problems are solved more rapidly when the relevant information is high in salience than when it is low in salience (Astor-Stetson e ta l , 1978; Cunningham and Odom, 1978; Katsuyama and Hoffarth, 1978; Odom et al, 1975; Overton and Jordan, 1971). Since older children generally solve problems more accurately and rapidly than younger children, many developmental psychologists have argued that older children have certain conceptual abilities which younger children do not have (e.g., Inhelder and Piaget, 1964). Odom and his colleagues entertain the alter native possibility that observed age differences in problem solving may be due largely to changes in perceptual sensitivity. They propose that older children, having had more experience with a variety of relations in the environment, are more sensitive to more relations than younger children. Consequently, in a given problem, the older child should be more likely than the younger child to evaluate more relations and thus should be more likely to solve the problem,
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
Perceptual Sensitivity and Conceptual Coordination
West/Odom/Aschkenasy
336
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
thereby demonstrating the presence of a particular conceptual ability. If this is the case, then age-related differences in problem solving would be due to differences in perceptual sensitivity rather than differences in conceptual pro cessing. Support for this view is provided by the evidence that younger children can solve problems that require the use of processes which conceptual-change theorists have attributed only to older children (Odom, 1978). For example, younger children solving coordination problems containing relevant information that is high in salience perform as well as older children solving conceptually identical problems containing relevant information that is low in salience (Cun ningham and Odom, 1978; Katsuyama and Hoffarth, 1978; Odom et al., 1975). There is considerable evidence from salience research with children as old as 12 years of age that the perceptual system continues to play an important role in problem solving throughout childhood. Nevertheless, there is a popular view expressed in the literature that the role of the perceptual system decreases with development while the role of the conceptual system increases (e.g., Wright and Vlietstra, 1975). That is, the younger child’s problem-solving performance is thought to be controlled by perceptual factors, and the older child’s perfor mance is thought to be controlled by the conceptual or logical requirements of a task. In contrast, the perceptual-salience position stresses the importance of the perceptual system throughout the life span. Given that the information in a problem must be perceived before it can be conceptually evaluated, such an emphasis would be tenable even in the absence of empirical support. It follows that at any point in the life span, the perceptual system’s sensitivity to informa tion should affect the probability that the information will be conceptually evaluated. To explore the influence of perceptual sensitivity on the development of coordination skills across the life span, a salience-assessment task and a coordina tion task were administered to younger and older adults as well as to children. The coordination task was quite different from those used in previous studies. Each subject was shown a series of items, each consisting o f three cards containing stimulus values representing two dimensions. The task required co ordination because the correct card (e.g., a red square) contained one dimen sional value that matched a second card (e.g., a red circle) and another dimen sional value that matched a value on a third card (e.g., a blue square). The task was presented in three phases which were designed to reduce the task require ments and increase the salience o f the relevant information over time. In phase I, single items were presented for the subject’s response. In phase II, the subject
Perceptual Sensitivity and Conceptual Coordination
337
was shown three items with the correct responses indicated, and was asked to respond to a fourth item. It was hoped that the change from a successive (phase I) to a simultaneous (phase II) presentation would reduce the memory requirements of the task. By looking at several solved items presented simul taneously, it was possible for the subject to eliminate certain inappropriate hypotheses before making a response to the unsolved item. In phase III, those subjects who did not solve the task in an earlier phase were told that the task required the coordination of certain unspecified information.
Method Subjects The participants in this study were 40 children, 10-14 years of age (mean = 11.9 years, SD = 0.74); 40 younger adults, 18-22 years of age (mean = 19.7 years, SD = 1.56); and 40 older adults, 60-78 years of age (mean = 69.4 years, SD = 4.93). Half of the individuals at each age level were male and half female. The children, 16 blacks and 24 whites, were enrolled in a predominantly middle-class school in Memphis, Tenn. The younger adults were from psychology classes at Vanderbilt University. The older adults were community-dwelling volunteers from senior centers and clubs in Memphis, Tenn. Three fourths of the older adults had completed a high school education, and their mean number of completed years of formal education was 13.1. All adult participants were white.
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
Materials Salience Assessment. The assessment task was used to determine the relative salience of the dimensions of form, color, and position. The task v/as a modified method of triads similar to the one used by Odom and Guzman (1972). Each salience-assessment item consisted of a 27.2 X 35 cm black card on which three 11.5 X 11.5 cm white cards were placed. Each white card had pictured on it a compound of four values, one from each of four stimulus dimensions. Color was represented by red, blue, green, yellow, orange, and purple. The forms were cross, square, circle, inverted T, triangle, and pentagon, each one no larger than 1.3 cm in diameter. Position values were represented by placement of stimuli either across the upper right to lower left diagonal, the upper horizontal edge, the right vertical edge, the vertical midline, the upper-left-to-lower-right diagonal, and the horizontal midlinc. The number values were 2, 3, 4, 5 ,6 , and 7. On all items, the number values were different on all three cards. There were four practice items. On the first practice item, the values of the three dimensions of form, color, and position were identical on two cards and differed on the third card. On each of the remaining practice items, one pair of cards had identical values on two dimensions while the values of the third dimension were different. The remaining card on these items differed from the other two cards on all dimensions. Each of the salience-assessment items contained three cards, two of which had identical values from one dimension and two of which had identical values from another dimension.
West/Odom/Aschkenasy
338
All three cards contained different values on the third dimension. Thus, two pairs of cards on each item had identical dimensional values. The three possible two-dimensional compari sons (form-color, form-position, color-position) were each presented six times. Problem-Solving Task. The problem-solving task was constructed from the same dimen sional values as those used during the salience-assessment procedure except that the value for number was always 3 and, therefore, number was held constant for all subjects. On each item of the task the subject was shown three cards side by side. Each of the 18 items was composed of a unique group of three cards. On each card was a compound stimulus com prised of values from several dimensions. There were three problems: (a) a form-color problem with form and color variable and relevant (position was held constant and was represented by placement of the stimuli on the horizontal midline); (b) a position-color problem where all the forms were squares, and (c) a form-position problem with all red stimuli as shown in figure 1. Each of the six values from the two relevant dimensions occurred three times in the 18 items. The correct card was randomly placed across trials with the restriction that it occurred an equal number of times in the left, right, and middle positions.
Fig. 1. Example of phase l (item d alone) and phase II (items a-d , with the correct card marked a on items a-c) of the coordination problem.
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
Procedure The experimenter began the salience-assessment task by showing the subject the first practice item, pointing to the two cards that were most similar and saying: They are more like each other than this card (pointing to the third card) which is different from these (pointing to the two most similar cards). On every page in this stack
Perceptual Sensitivity and Conceptual Coordination
339
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
there will always be cards that are more alike. You look at the cards on each page and point to the cards that are more alike. The experimenter then administered the next three practice items followed, without interruption, by the 18 salience-assessment items. The subject’s choice for each item was recorded. The frequency of choices based on each dimension determined the subject’s salience hierarchy, in which the most salient dimension was A (the most frequently chosen) and the least salient dimension was C. If a subject picked two dimensions an equal number of times, those two dimensions were designated an AB tie or a BC tie. After the experimenter determined the individual’s salience hierarchy, the problem solving task was administered. 20 subjects at each age level were assigned to two groups one group was given a task where the subjects’ A and B dimensions were variable and relevant and the C dimension was constant (the AB condition) while the other group received a task where the subjects' B andC dimensions were variable and relevant while the A dimension was held constant (the BC condition). If a subject had an AB tie on the salience-assessment task, he was assigned to the AB condition and if he had a BC tie he was assigned to the BC condition. All other subjects were randomly assigned to the two condi tions. Since the placement of the cards in the array (left, right, or middle) was variable irrelevant information, subjects were told in the instructions that placement was irrelevant. Thus, there were only two variable relevant dimensions in each problem. As the experimenter gave the task instructions, she showed the subject a sample item. The three cards in the sample items contained human cartoon characters taken from maga zines. These characters varied in color, size, shape, apparent age, type of clothing, direction the character faced, etc. The subjects were given the following directions: Today I am going to give you a learning task. I am going to show you a large number of pictures, three at a time, arranged like this (pointing to the sample item). Each time I show you three pictures, one of the three pictures will be the correct picture, according to a rule. You will need to discover the rule. You will have to guess at first, but later you will discover the rule and you will be able to pick the correct picture all the time. The experimenter then gave an example of a rule for the sample item, and continued with the instructions: The rule has nothing to do with where the correct picture occurs in the set. Now each time I show you three pictures, you point to the picture which you think is the correct picture. Then I will tell you which picture is correct. Pay close attention to all the pictures so you can figure out the rule which tells you which pictures are correct. Once you know the rule, you will always be able to pick the correct picture in each set of three pictures. The solution rule for the concept identification problem was this: pick the card which has the same value on one dimension (B) as another card in the set of three and the same value on the second dimension (A or C) as the remaining card. For example, in the formposition problem in figure 1, item d, the correct card (3) matched the form value in card 2 and matched the position value in card 1. In phase I of the task, each of the 18 items was presented one at a time for the subject’s response. After each response, the experimenter indicated the correct card by pointing to it and saying ‘Yes (No), this is the correct picture’. If the subject did not reach the solution criterion (six consecutive correct responses and an explanation of the rule) during the initial presentation of the 18 items, the second phase of the task began.
West/Odom/Aschkenasy
340
In phase II the experimenter presented the original 18 items four at a time. The subject was told: Now 1 am going to show you four sets of pictures at the same time. I will give you the answer for the first three, then you try and pick the correct picture for the fourth set. The rule has to do with each set of three pictures by itself I’m only doing this so you can see several examples at one time. Remember, the rule has nothing to do with whether the correct picture is on the left, right, or in the middle. The first four items were placed before the subject, and the experimenter put a token on the correct cards for the first three items, saying ‘this is the correct picture here’ (see fig. 1 for an example of the form-position problem). The subject was then asked to pick the correct card for the fourth item. After the response and feedback, the experimenter picked up items 1 -4 and similarly placed the next four items in front of the subject. This proce dure continued for six sets of four items, or until the subject reached a criterion of three consecutive correct responses and an explanation of the correct rule. If the subject did not solve the problem with this procedure, phase III began. In phase III the experimenter told the subject, ‘The rule has to do with how the pictures are alike. So look closely to see how the pictures are alike.’ Three more sets of four items were presented as in phase II, and then the experimenter provided another hint: ‘There are two ways that the pictures are like each other. The rule has to do with two ways that the pictures are alike.’ If the subject did not reach criterion after three more sets of four items were presented, the experimenter provided more hints: ‘How is this picture (pointing to the last correct card) like the other two pictures here?' This final procedure was necessary in four cases for the sixth graders and four cases for the older adults. One sixth grader and one older adult in the BC condition did not solve the problem after six more sets of four items were presented and these two subjects were eliminated from the data analysis. There were three dependent measures recorded for each subject: (a) the number of trials to criterion; (b) the percentage of errors, i.e., the number of errors divided by the number of trials to criterion, and (c) the type of errors, whether errors based on form, color, or position. For example, a form error occurred when the subject in the form problem picked the incorrect card which matched the correct card on form.
Results
Downloaded by: UCL 144.82.238.225 - 1/11/2018 5:03:12 PM
The mean number of trials to criterion and the mean percentage of errors for each age group in the AB and BC problems are presented in table I. A 2 (condition) X 3 (age) analysis of variance was performed on the dependent variable of trials to criterion. There was a main effect for condition, F (l, 114) = 16.8, p
