Hum. Dev. 21: 302-315 (1978)

Methodological Inconsistencies in the Measurement of Spatial Perspective Taking Ability: A Cause for Concern Lawrence A. Fehr University of New Orleans, New Orleans, La.

Key Words. Childhood development • Cognitive development • Egocentrism • Perspective-taking • Piaget • Spatial organization • Spatial perception Abstract. A review of the literature pertaining to spatial perspective-taking is presented. Its purpose is to account for the inconsistent findings that have plagued this area of research by focusing on the methodological differences between studies. The conclusions reached are threefold. First, methodological differences are at least in part responsible for the widely discrepant results that have been reported. Second, the amount of variation due to each methodological concern cannot be quantified precisely due to the problem of confounding. Third, a viable measure from which the development of perspective ability can legitimately be inferred cannot be conclusively identified within the perspective-taking literature.

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

A crucial area within the study of cognitive development involves the development of spatial abilities. The work of Piaget and Inhelder (1956) has done a great deal to establish the importance of studying these abilities. It has also led to a great deal of controversy. An analysis of research performed in this area leads one to conclude that a focal point of much of the controversy has been the nature of the development of perspective ability or the ability to coordinate one’s own point of view with that of another person or object. Piaget and Inhelder have referred to this as the ability to coordinate perspectives. They studied this ability by asking children to indicate the position in which a doll would have to be sitting in order to have taken a given picture of a mountain scene. Each child was also asked to indicate which of a set of ten pictures looked like what the doll could see from a given position. Finally, each child was asked

Spatial Perspective-Taking

303

to construct three-dimensional representations of the mountain scene in accord with given viewing positions of the doll. The results reported by Piaget and Inhelder (1956) indicated that until the age of 4 -5 years, the children were unable to comprehend the nature of a Coordination of Perspectives task. The next developmental stage which was found to last from the age of 5 -7 years was typified by totally egocentric response patterns. That is, these children were found to believe that theirs was the only possible point of view. From the age of 7 years onward, there was a progressive decrease in egocentrism, until at about the age of 9 -1 0 years the children were able to coordinate perspectives with a high degree of accuracy. It is a well-established fact {Lewis and Fishbein, 1969; Fishbein et al., 1972; Eiser, 1976; Cox, 1975, 1977) that a wide variety of task differences have contributed to inconsistencies between the findings of Piaget and Inhelder (1956) and more recent research. However, no attempt has yet been made to meaningfully summarize the findings of these and other relevant studies. There­ fore, it is the purpose of this review to account for the various inconsistencies that exist within the perspective literature by examining the potential contribu­ tory effects of various methodological differences between these studies. The methodological constraints to be considered include the number and type of stimuli depicted in the spatial array, the orientation(s) of the other observer (hereafter referred to as the ‘other’) regarding that of the subject, the nature of the task performed by the subject, the number and type of choice stimuli, and the animate or inanimate nature of the ‘other’ (doll or experimenter) with whom or which the subjects must coordinate their perspectives.

An early attempt at replicating the findings of Piaget and Inhelder relative to stages in perspective development was conducted by Dodwell (1963) who used stimuli which were similar to those employed in the Piaget and Inhelder study. Dodwell concurred with Piaget and Inhelder relative to developmental trends in error rates. However, the percentage of subjects’ errors which were egocentric was not found to be related to the overall performance of the subjects. Even though it was not very critical of the methodology employed by Piaget and Inhelder, Dodwell’s study was useful in that it did show that Piaget and Inhelder’s results were replicable if considered within die relatively strict context of their own methodology.

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

Number and Type of Stimuli in the Spatial Array

304

Perhaps the most thorough attempt to replicate the findings of the Piaget and Inhelder study was performed by Laurendeau and Pinard (1970) who substituted three cones for the cardboard mountains which Piaget and Inhelder used as stimuli. There was a discrepancy between the error rates reported in this study and those reported by Piaget and Inhelder. Laurendeau and Pinard found that children as old as 12 years of age had error rates as high as 40%. The late age of mastery of the Coordination of Perspectives reported in this study and elsewhere (Garner and Plant, 1972) has been attributed to several factors. The first is that the stimuli employed in these studies were unfamiliar to the subjects (Eiser, 1974). That is, subjects have been found to perform better with actual mountain scenes (with animals, houses, etc.) than with cones and other shapes that lack an obvious referent for the subject. A second factor which is closely related to familiarity is differentiation. Not only are cows, houses, toys, etc., more familiar to children than cones, they also have more distinctive features. The importance of this factor has been supported experimentally be Shlechter (1977) who found that perspective-taking ability is more accurate with arrays consisting of houses that possess a variety of distinctive features than with houses that lack such differentiation. The results secured by Piaget and Inhelder (1956), Dodwell (1963), and Laurendeau and Pinard (1970) were relatively similar. However, so were their stimuli. Recent research has demonstrated that more substantial stimulus changes have yielded results that differ dramatically from those already re­ ported. It can be recalled from the Piaget and Inhelder study that the ability to perform a coordination of perspectives task asymptotes at about 9 -1 0 years of age. This claim has been sharply rebutted by Borke (1975). Using a display consisting of a toy sailboat on a lake, a miniature horse and a cow, and a model of a house, Borke asked children to move a turntable in a manner that would enable a scene to look to the child the way that a duplicate scene looked to a fictional character who drove his motorcycle to various points relative to the scene. Borke found that children manifested little egocentrism beyond the age of four years on this task. This early decline in spatial egocentrism has been supported by the findings of several other studies {Shantz and Watson, 1970; Masangkay et al., 1974). Thus, Borke (1975) has concluded that spatial egocen­ trism is not a necessity for the preoperational child. Rather, it is a response pattern that is dictated by requiring the child to perform a task that is too difficult for his/her level of cognitive development. This claim has also been substantiated by Aebli (1967) who found that the frequency of egocentric responding is partially a function o f whether or not the subject is asked to select

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

Fehr

305

his/her own view of the stimuli during the testing. When this step was included in the methodology, Aebli found that egocentric responses occurred more frequently. It is clear that subjects’ performances on perspective-taking tasks are not totally dependent upon the types of stimuli that are used. However, the tremendous discrepancies between the results of Laurendeau and Pinard (1970) and Borke (1975) must be accounted for. Within the context of the present issue, which is type of stimuli, the two key words appear to be familiarity and differentiation. The relevance of these issues has already been alluded to as espoused by Eiser (1974) and Shlechter (1977), respectively. Although there is no clear-cut solution to this problem, it does appear that the situation can be summarized in the following manner. The findings of Borke (1975) have shown that it is possible to develop a stimulus situation in which children as young as 4 years of age can respond appropriately. However, it does not necessarily follow that one must conclude that perspective ability is fully developed by 4 years of age based upon the results of studies employing simple, familiar stimuli. If one has mastered perspective tasks, he/she should be able to appropriately deal with relatively complex stimuli. On the other hand, it is also possible that the stimuli employed by Laurendeau and Pinard (1970) have fostered error rates that are artificially high. The effect of the number of objects in a spatial array on the ability of subjects to perform a Coordination of Perspectives task has been examined in a number of studies (Flavell et al., 1968; Fishbein ei al., 1972;Brodzinsky et al., 1972; Minnigerode and Carey, 1974;/toy, 1974; Borke, 1975). The contradic­ tions that exist relative to this variable are typical of those that abound in tire perspective literature. Comparing the ability of children to perform a Coordina­ tion of Perspectives task using 1 vs. 3 toys as stimuli, Fishbein et al. (1972) found that subjects ranging in age from 3.5 to 9.5 years did significantly better on the one toy than on the three toy condition. This finding is consistent with those reported by Flavell et al. (1968) and Hoy (1974). However, no such effect was reported by Brodzinsky et al. (1972), Minnigerode and Carey (1974), or Borke (1975). This discrepancy can be attributed to numerous factors, several of which (rotation, number of choice stimuli, etc.) will be considered shortly. For the present, an analysis of these differences will be restricted to a consideration of the relationship between the number of objects in an array and task difficul­ ty. There is no guarantee that increasing the number of objects in a spatial array also increases task difficulty. This point can best be exemplified by the findings of Fehr and Fishbein (1976). This study compared the ability of subjects to

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

Spatial Perspective-Taking

Fehr

306

recognize correct spatial representations of three and four object arrays. The two arrays were comprised of three identical objects while the fourth object was correctly arrayed on all of the choice stimuli (the subjects were informed of this). The presence of the additional object resulted in superior performances across age groups. The point of the above analysis based on the results of Fehr and Fishbein (1976) is that the effect of increasing the number of objects in an array cannot be assumed to be the same across the variety of tasks and stimulus arrays employed in perspective studies. However, returning to the basic issue of the development of perspective ability, statements in this regard should be based upon research using multiple object arrays. Single object arrays provide subjects with situations in which they may often 'memorize’ different perspectives of the object. This can become especially pronounced if the object is a simple one. As has previously been stated when the issues of familiarity and differentiation were considered, developmental trends in perspective ability should not be dictated by the performance of subjects on oversimplified tasks.

Orientation of the 'Other'

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

A critical problem that has led to a great deal of difficulty in comparing the results of perspective studies is differences in the number and type of orienta­ tions of the ‘other’ relative to that of the subject. In some studies (Hoy, 1974; Sliantz and Watson, 1971; Minnigerode and Carey, 1974), the ‘other’ is located in only one rotated position relative to the orientation of the subject. This position is directly opposite the subject or at a 180° rotation relative to the orientation of the subject. Other studies (Dodwell, 1963; Brodzinsky et al., 1972; Fishbein et al, 1972) have employed three rotations of the ‘other’ (90°, 180°, 270°). However, in these studies, no attempt was made to determine the differential effects o f having the ‘other’ located across (180°) versus to the side (90°, 270°) of the subject. Additional studies have examined the child’s ability to coordinate perspectives by locating the ‘other’ at positions involving one or more 45° rotations (Flavell et al., 1968; Laurendeau and Pimrd, 1970; Coie et al., 1973; Rubin, 1974; Walker and Collin, 1977). Based upon the results reported in the studies cited above, it can be stated that experiments which have utilized 45° rotations of the position of the ‘other’ have yielded the poorest performances. For example, on the three cylinders task,

Spatial Perspective-Taking

307

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

Flavell et al. (1968) found that errors were common for subjects as old as 16 years of age and that egocentrism declined at about 12 years of age. In addition, as was alluded to earlier, Laurendeau and Pinard (1970) found error rates as high as 40% for subjects as old as 12 years of age. These findings can be contrasted with those of several 90° rotations studies (Dodwell, 1963; Fishbein et al, 1972) which found egocentrism to decline between the ages of 7.5 and 9.0 years. Unfortunately, other task and stimulus differences between these studies render it difficult to make any unequivocal statement relative to the effect of orientation of the ‘other’ on subjects’ abilities to perform a Coordination of Perspectives task. Several attempts have been made to determine tire relative difficulty that children have in dealing with different orientations of the ‘other’. The findings of these studies (Nigl and Fishbein, 1914,Eiser, 1974; Walker and Gollin, 1977; Cox, 1977) have contradicted one another. Nigl and Fishbein (1974) and Cox (1977) both examined successive 90° rotations of the ‘other’ (90°, 180°, 270°) and compared the results that were secured in the across (180°) and the side (90°, 270°) conditions. In both of these studies the across condition was more difficult than the side conditions. Eiser (1974) compared across, side, and corner (45°, 135°, 225°, 315°) positions of the ‘other’. The results were that the across condition was judged as the easiest of the three followed by the comer conditions. Finally, Walker and Gollin (1977) found comer views to be the most difficult. Although these findings contradict one another, the most convincing argument was made by Cox (1977). In claiming that the opposite view was the most difficult, she pointed out that in Eiser's study, masking of a small object by a large one occurred in the across condition. As has been pointed out by Flavell et al. (1968), such masking tends to make a task easier for the subject. Even if the findings of the above studies had been consistent with one another, the effects of different orientations of the ‘other’ still would not have been adequately accounted for. For example, even if the across condition had been consistently judged as being more difficult than the side conditions, it would not necessarily follow that a task employing the 180° condition only would be more difficult than a task involving successive 90° rotations. The studies cited compared the effects of different rotations of the ‘other’ in a within subjects design. A between subjects design in which one group was exposed to the across condition only, another group to the side conditions only, and a third group to both the across and side conditions would appear to be the only way to accurately determine the role played by orientation of the ‘other’ in a Coordination of Perspectives task. In addition, this design could obviously be

Fehr

308

extended to include a variety o f other rotations of the position occupied by the ‘other’. Returning once again to the basic question of how perspective ability develops, it is clear that subjects should be exposed to a variety of positions of the ‘other’. Only in this manner can an individual’s perspective ability manifest flexibility which should be a criterion in the establishment of overall trends in the development of perspective ability.

Numerous minor task differences can be found to exist across studies that have examined the development of perspective ability. Most research methodolo­ gies used in this area require that tire subject perform an identification task in which he/she attempts to identify a correct photograph or other spatial repre­ sentation of an object array that coincides with various perspectives of the ‘other’. There is also a much less frequently used methodology wliich requires the subject to construct or arrange a three-dimensional array in a manner that is consistent with a given perspective of the ‘other’. An obvious trend that can be noted in the literature is that the number of overall as well as egocentric errors per subject is greater on construction than on identification tasks. The notion of a construction task is based on the third part of the Piaget and Inhelder (1956) methodology that has purposely been ignored by most researchers. As Laurendeau and Pinard (1970, p. 320) point out: ‘It was decided to discard this third technique, wliich was too difficult for the children, wliich they very soon lost interest in, and whose results lend themselves to objective scoring only with great difficulty.’ This is consonant with the views of most perspective researchers. Nevertheless, this type of task was resurrected in kind by Flavell et al. (1968) in their three cylinders task which required subjects to arrange a set of materials in accord with the perspective of the ‘other’. As has previously been stated, egocentrism did not decline until 12 years of age on this task wliich was also employed by Rubin (1973, 1974) with similar results. Subsequently, the greater difficulty of a construction or arrangement as com­ pared to an identification task was substantiated by Hoy (1974). It would appear to be intuitively obvious that it is illogical to evaluate the development of perspective ability based upon the results of a construction or arrangement task. The perceptual-motor skills of the child become prominent in

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

Type of Task Performed by the Subject

309

these tasks which apparently represent an opposite extreme in difficulty to tire methodology employed by Borke (1975). While the latter was purported to measure something less inclusive than perspective ability, the former measures something in addition to perspective ability. A task distinction which is relevant to our understanding of the develop­ ment of perspective ability can be found to exist within the large number of studies whose basic methodology is one of identification. It has been noted by Eiser (1974) that this methodology can be subdivided into tasks which require of the subject either recognitory or inferential knowledge. In the former, the subject is permitted to view the spatial array from all possible perspectives before beginning the test trials. As a result, on the test trials, the subject need only recognize a perspective that has already been viewed during tire familiariza­ tion procedure. In the remaining studies, the subject is not permitted to view other perspectives during the familiarization procedure. The subject must there­ fore make use of inferential knowledge in making spatial judgments. Eiser (1974) hypothesized and subsequently found that subjects performing a recogni­ tion-perspective task made significantly fewer errors than subjects performing an inferential-perspective task. It is difficult to evaluate Eiser's hypothesis pertaining to the relative difficul­ ty of various perspective procedures in light of previous findings due to method­ ological differences across studies. However, the trends that can be noted are interesting. In studies employing the recognition procedure (Houssidas, 1965; Fishbein etal., 1972; Brodzinsky et al., 1972; Coie et al., 1973) egocentrism was found to decline at a mean age of approximately 8.5 years while in studies employing the inferential procedure (Piaget and Inhelder, 1956; Dodwell, 1963; Laurendeau and Pinard, 1970; Larsen arid Abravenel, 1972) the mean age at which egocentrism declined was approximately 9.5 years. Perhaps more than any other factor, this issue of recognition versus inferential procedure exemplifies that criticisms of the results of the original Piaget and Inhelder study have in many cases been based on replications that not only were not ‘true’ replications but which may not even have been replications ‘in kind’. In terms of the justification for using one of the two types of tasks mentioned above, it would appear that the use of an inferential task can be more easily justified. This conclusion is based on the premise that if the practice and test trials are similar in a recognition task, subjects’ performances may in part be determined by their ability to recollect recently viewed perspectives. This possibility could lead to a variety of methodological problems, particularly if one is concerned with the processes involved in making perspective judgments.

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

Spatial Perspective-Taking

Fehr

310

A factor that has often been ignored in the perspective literature involves variations in tire choice stimuli. Three aspects of this issue are worthy of consideration. These aspects are the two- or three-dimensional nature of the choice stimuli, the number o f choice stimuli employed on each trial, and finally, the types of distortions that are depicted in the incorrect choice stimuli. Relative to the issue of the dimensionality of the choice stimuli, most studies {Piaget and Inhelder, 1956, Dodwell, 1963; Houssidas, 1965; Laurendeau and Pirnrd, 1970; Shantz and Watson, 1971; Nigl and Fishbein, 1973) have used as stimuli two-dimensional photographs of the stimulus arrays. However, one recent study (Nigl and Fishbein, 1974) did compare the ability of children to coordinate perspectives using two- and three-dimensional choice stimuli. The findings of this study were that children were better able to coordinate perspec­ tives when three- rather than two-dimensional choice stimuli were used. This leads one to reexamine the question, what should be the composition of a viable measure of perspective ability? It is obvious that two-dimensional choice stimuli are easier to construct and more manageable than three-dimensional choice stimuli. However, is it not possible that studies employing three-dimensional arrays and two-dimensional choice stimuli are measuring some other skill in addition to perspective ability? A solution to this problem would appear to be to compare the performances o f subjects divided into four groups. The groups could be divided according to the dimensionality of the original array as well as the dimensionality of the choice stimuli. Tliis would enable a determination to be made relative to the interaction between the dimensionality of the original array and the dimensionality of the choice stimuli. An analysis of the data from such an experiment would better enable researchers to determine if the dimen­ sionality of the choice stimuli is a variable that must be considered in attempting to construct a viable measure of perspective ability. A consideration of the number of choice stimuli used in perspective tasks leads to findings that are somewhat inconsistent. The number of choice stimuli per trial has in some cases been as few as four {Houssidas, 1965) or five {Laurendeau and Pirnrd, 1970; Shantz and Watson, 1971), and in other cases it has been as many as ten {Piaget and Inhelder, 1956; Dodwell, 1963) or 20 {Miller, 1967). The results of these studies are inconclusive in that high error rates have been found using both few {Laurendeau and Pirnrd, 1970) and many {Miller, 1967) choice stimuli. However, the results are consistent in that studies

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

Number and Type of Choice Stimuli

Spatial Perspective-Taking

311

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

that have employed ten or more choice stimuli per trial have universally secured error rates that are above the mean for all perspective studies. The fact that there is a relationship between the number of choice stimuli used on each trial and subjects’ error rates should not be surprising if for no other reason than that the probability of making a correct choice (by chance) given five alternatives is 0.2 while for ten alternatives it is 0.1. Just as this factor affects error rates, it could also affect the percentage of subjects’ errors that are egocentric. It is logical that one is more likely to make an egocentric error if there are four or five incorrect choices one of which is egocentric, than if there are nine incorrect choices with only one representing an egocentric error. This accentuates the point that in examining egocentrism and error rates, future research should focus upon increases and decreases across age and conditions rather than actual percentages which must be stimulus specific. A factor that is closely related to the number of choice stimuli used in perspective studies is the types of errors that are depicted in the incorrect choice stimuli. The most common pattern that exists is to present a series of rotations and one or more impossible views of the array (Laurendeau and Pinard, 1970; Shantz and Watson, 1971; Coie et al., 1973;£7ser, 1974). However, these studies have failed to determine which rotations of the choice stimuli are most likely to lead to incorrect responses on the part of the subjects. In order to better answer this question, the geography literature must be considered. In a study which has been ignored in the perspective literature, Miller (1967) presented children with 20 choice stimuli depicting successive 18° rotations of a spatial array. Miller found that the most common errors made by subjects involved choosing photographs which represented 18° rotations relative to the correct choice. These results provide an interesting point for discussion in that rotations of less than 45 ° are typically ignored in the construction of choice stimuli. This could account for the relatively low error rates that have been recorded in some studies. However, the question that must be asked at this point is, how accurate must a child be? Is it unreasonable to except a child to be accurate to the nearest 18°? The best solution to this problem would appear to be to look at both fine tuned accuracy as well as approximate accuracy. Methodologically, this could be handled by presenting a child with a series of rotations and analyzing the data two ways. In the first analysis, only precisely correct responses would be scored as correct while a second analysis could be based on a more lenient scoring procedure. The findings that have been noted relative to the choice stimuli used in perspective tasks indicate that this factor has been inappropriately ignored in the

Fehr

312

literature. Studies have used a variety of choice stimuli without in most cases demonstrating any obvious rationale for their selection. Moreover, most studies have failed to analyze which type of spatial representations are most likely to be judged as being correct ones. When they have been performed, such analyses have typically been limited to a consideration of egocentric versus nonegocentric errors. The results reported by Miller (1967) have provided proof that the selection of choice stimuli in a perspective study affects the results of the study. More importantly, this selection affects the nature of statements that are made relative to the how and the when of the development of perspective ability.

Animate vs. Inanimate 'Other'

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

The final methodological issue to be considered pertains to tire nature of the ‘other’ in perspective tasks. Reiterating what has been stated previously in this regard, the ‘other’ is the person or object with whom (which) the child must coordinate his/her perspectives. In the original Piaget and Inhelder study, the ‘other’ was a doll that was placed at various points on the perimeter of the spatial array. Following the lead of Piaget and Inhelder, many perspective studies (Dodwell, 1963; Houssidas, 1965; Laurendeau and Pinard, 1970) continued to use a doll as the ‘other’. The importance of this methodological consideration was alluded to by Fishbein et al. (1972) who used a human experimenter as the ‘other’. They pointed out that asking a child to conceive of the viewpoint of a doll is a hypothetical problem because a doll can neither see nor take a photograph. Thus, if a child has difficulty with hypothetical problems (as preoperational and concrete operational children are purported to have accord­ ing to Piagetian theory) they might be able to conceive of the viewpoint of an experimenter but not a doll. The issue of the role played by an animate versus an inanimate ‘other’ has recently been empirically tested by Cox (1975). This study compared the ability of children to coordinate perspectives with a doll and a person. The results of the study were that egocentric and overall error rates were higher for the doll than the experimenter group. An analysis of the role played by the nature of the ‘other’ in perspective tasks leads one to conclude that researchers who wish to measure perspective ability should employ a person as the ‘other’ because a doll cannot have a perspective. As was the case with several other seemingly harmless manipulations that have already been considered, using a doll as the ‘other’ appears to have

Spatial Perspective-Taking

313

complicated the perspective task to the extent that what is being measured is something more than perspective ability. Rather than replicating that aspect of Piaget and Inhelder'% methodology, researchers should have immediately im­ proved upon it. Determining the reliability of Piaget and lnhelder'% findings is not nearly as noble a task as determining trends in the development of perspec­ tive ability. As can be surmised from the work of Fishbein et al. (1972) and Cox (1975), Piaget and lnhelder'% methodology of using a doll as the ‘other’ may have fostered artificially high egocentric and overall error rates by asking preoperational and concrete operational children to deal with tire hypothetical. Although tire basis of the claim is different, the notion that some of the trends reported by Piaget and Inhelder (1956) were based upon inappropriate procedur­ al tecluriques has been offered on several occasions (Aebli, 1967; Borke, 1975).

Summary: Implications of Methodological Inconsistencies on an Understanding of the Development of Perspective Ability

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

A number of methodological differences have been noted between the numerous studies that purport to measure perspective ability. Unfortunately, much of the research performed in this area has concentrated on an examination of various aspects of Piaget and lnhelder'% findings on the subject. In order to accurately determine developmental trends in perspective ability, studies must be designed that focus upon the development of perspective ability only. The results that have been secured from perspective studies have been inconclusive from a theoretical point of view. The sundry methodological variations that exist have led researchers to conclude that spatial egocentrism declines either as early as 4 years of age (Shantz and Watson, 1970) or as late as 12 years of age (Flavell et al., 1968). Methodological inconsistencies have also led researchers to conclude that error rates on perspective tasks become minute either as early as 4 years of age {Borke, 1975) or in excess of 12 years of age (Laurendeau and Pinard, 1970). The only claims that can be made with some degree of certainty are that overall accuracy increases with age and egocentrism eventually declines with age. However, these claims are akin to stating that a 6-year-old boy will be taller when he is 12 years old. Based on the number of inconsistencies that have been found relative to previous research, it will take several additional experiments to develop a viable measure of perspective ability. However, the evidence that is currently available indicates that a good test of perspective ability only should be constructed

Fehr

314

within the following guidelines. (1) A person should serve as the other observer. (2) The other observer should occupy a wide variety of positions. (3) Inferential rather than recognition or construction tasks should be used. (4) The spatial stimuli should consist of arrays of more than one object. At one time or another, one or more studies has either suggested or performed each of the above variations in perspective methodology. However, at no time have these factors been considered simultaneously. When this is experi­ mentally accomplished, we should be a step closer to understanding how perspective ability develops and also when spatial egocentrism declines as measured by a ‘true’ perspective task.

References

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

Acbli, H.: Egocentrism. Piaget not a phase of mental development but a substitute solution for an insoluble task. Pedagog. Eur. 3: 97-103 (1967). Borke, H.: Piaget’s mountains revisited. Changes in the egocentric landscape. Devi Psychol. 11: 240-243 (1975). Brodzinsky, D.M.; Jackson, J.P., and Overton, W.F.: Effects of perceptual shielding in the development of spatial perspectives. Child Dev. 43: 1041-1046 (1972). Coie, T.; Costanzo, P., and Farnhill, D.: Specific transitions in the development of spatial perspective taking ability. Devi Psychol. 9: 167-177 (1973). Cox, M.V.: The other observer in a perspective task. Br. J. educ. Psychol. 45: 83-85 (1975). Cox, M.V.: Perspective ability. The relative difficulty of the other observer’s viewpoints. J. exp. Child Psychol. 24: 254- 259 (1977). Dodwell. P.C.: Children’s understanding of spatial concepts. Can. J. Psychol. 17: 141-161 (1963). Eiser, C.: Recognition and inference in the coordination of perspectives. Br. J. educ. Psychol. 44: 309-312 (1974). Eiser, C.: Questions children ask about spatial arrays. An analysis of the processes involved in coordinating perspectives. Br. J. educ. Psychol. 46: 203-211 (1976). Fehr, L.A. and Fishbein, H.D.: The effects of an explicit landmark on spatial judgments; in Suedfeld and Russell, The behavioral basis of design (Dowden, Hutchinson & Ross, Stroudsburg 1976). Fishbein, H.D.; Lewis, S., and Keiffer, K.: Children’s understanding of spatial relations. Coordination of perspectives. Devi Psychol. 7: 21-33 (1972). Flavell, J.H.; Botkin, P.T.; Fry, C.L.; Wright, J.W., and Jarvis, P.E.: The development of communication and role taking skills in children (Wiley, New York 1968). Garner, J. and Plant, E.L.: On the measurement of egocentrism. A replication and extension of Aebli’s findings. Br. J. educ. Psychol. 42: 79-83 (1972). Houssidas, L.: Coordination of perspectives in children. Arch. ges. Psychol. 117: 319-326 (1965).

Spatial Perspective-Taking

315

Hoy, E.A.: Predicting another’s visual perspectives. A unitary skill? Devi Psychol. 10: 462 (1974). Larsen, G.Y. and Abravenel, E.: An examination of the developmental relations between certain spatial tasks. Merr. Pal. Q. 18: 39-51 (1972). Laurendeau, M. and Pinard, A.: The development of the concept of space in the child (International Universities Press, New York 1970). Lewis, S. and Fishbein, H.D.: Space perception in children. A discontinuation of Piaget’s developmental hypothesis. Psychonomic Soc. Meet. St. Louis 1969. Masangkay, Z.S.; McCluskey, C.W.; Sims-Knight, J.; Vaughn, B.E., and Flavell, J.H.: The early development of inferences about the visual percepts of others. Child Dev. 45: 357-366 (1974). Miller, J.W.: Measuring perspective ability. J. Geog. 66: 167-171 (1967). Minnigerode, F.A. and Carey, R.N.: Development of mechanisms underlying spatial perspec­ tives. Child Dev. 45: 496-498 (1974). Nigl, A.J. and Fishbein, H.D.: Children’s ability to coordinate perspectives. Cognitive factors; in Preiser, Environmental design research (Dowden, Hutchinson & Ross, Stroudsburg 1973). Nigl, A.J. and Fishbein, H.D.: Perception and conception in coordination of perspectives. Devi Psychol. 10: 858-866 (1974). Piaget, J. and Inhelder, B.: The child’s conception of space (Routiedgc, London 1956). Rubin, K.H.: Egocentrism in childhood. A unitary construct. Child Dev. 44: 102-110 (1973). Rubin, K.H.; The relationship between spatial and communicative egocentrism in children and young and old adults. J. gen. Psychol. 125: 295-301 (1974). Shantz, C.W. and Watson, J.S.: Assessment of spatial egocentrism through expectancy violation. Psychon. Sci. 19: 93-94 (1970). Shantz, C.W. and Watson, J.S.: Spatial abilities and spatial egocentrism in the young child. Child Dev. 42: 171-181 (1971). Shlechter, T.M.; Children’s spatial coordination and the influences of environmental dif­ ferentiation. Soc. Res. Child Dev. Meet., New Orleans 1977. Walker, L.D. and Gollin, E.S.: Perspective role-taking in young children. J. exp. Child Psychol. 24: 343-357 (1977).

Downloaded by: Kings's College London 137.73.144.138 - 11/11/2017 4:21:38 PM

Lawrence A. Fehr, Department of Psychology, University of New Orleans, New Orleans, LA 70122 (USA)

Methodological inconsistencies in the measurement of spatial perspective taking ability: a cause for concern.

Hum. Dev. 21: 302-315 (1978) Methodological Inconsistencies in the Measurement of Spatial Perspective Taking Ability: A Cause for Concern Lawrence A...
1MB Sizes 0 Downloads 0 Views