H H l H DiSSiPATEOR GF FUNCTIONAL FIXEDNESS BY MEANS 9F CQNCEPTUAL {ERO’dPiNG “(ASKS \ M \ I]; I! ma (ADM 00V Thesis for the Degree of Ph. D. MICHIGAN STAEE UNWERSITY ROBERT H. TERBDRG 1988 THEHG A, ‘ Mm UZ'IIVL may This is to certify that the thesis entitled DISSIPATION OF FUNCTIONAL FIXEDNESS BY MEANS OF CONCEPTUAL GROUPING TASKS presented by Robert H. Terborg has been accepted towards fulfillment of the requirements for Ph.D. Psychology degree in W / 15/4” [é/‘flL/V . Major ’pro egalit- (r Date wéfj/LIZ/fl: /Zc// 0-169 i 7 BINDING BY ““ ms & sow L am mom mc. ABSTRACT DISSIPATION OF FUNCTIONAL FIXEDNESS BY MEANS OF CONCEPTUAL GROUPING TASKS by Robert H. Terborg The purpose of this experiment was to examine the effect of conceptual grouping according to unusual func- tions on functional fixedness, using Glucksberg's circuit problem. Prior to being asked to solve the circuit prob- lem §§ were randomly divided into five equal groups. Each group received practice on a conceptual grouping task. Subjects in the control group (C) were asked to group ob— jects on the basis of a common attribute. All four experi- mental groups (El' E2, E3, E4) were asked to group objects on the basis of unusual functions. The experimental groups differed in the extent to which they were asked to make groupings which stressed an electrical function (this was the critical function of the circuit problem which they were asked to solve after completing the conceptual group- ing task). 'On the basis of S-R theory and Gestalt theory it was hypothesized that: Robert H. Terborg (1) Practice with conceptual grouping tasks which stress only unusual use of objects (i.e., one which does not stress the critical function) will not reduce functional fixedness. (2) Practice with conceptual grouping tasks which stress unusual use of objects for the critical function should reduce the functional fixedness in the circuit problem. Furthermore the amount of time needed to solve the problem should be a direct function of the amount of practice with the critical function. The results supported the first hypothesis. The second hypothesis was not confirmed but there was a strong trend in the predicted direction. The results were con- sistent with either S-R or Gestalt theory. Approved Donald M. Jo son Chairman Date DISSIPATION OF FUNCTIONAL FIXEDNESS BY MEANS OF CONCEPTUAL GROUPING TASKS BY ,\J \‘v’ , u Robert HilTerborg A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1968 To my wonderful parents. ii ACKNOWLEDGMENTS The author wishes to express his sincere apprecia- tion to Dr. Donald M. Johnson, the Committee Chairman, for his interest, encouragement, and helpful suggestions throughout the development of this manuscript. The author also wishes to thank the members of the committee, Dr. Gordon Wood, Dr. Joe L. Byers and Dr. William Stellwagen for their reading and criticism of the manuscript. iii INTRODUCTION . METHOD . O O 0 Subjects TABLE Experimental Task . . Procedure Measures RESULTS . . . DISCUSSION . . SUMMARY . . . REFERENCES . . OF CONTENTS iv Page 11 11 ll 12 16 18 22 27 29 LI ST OF TABLES Table Page 1. Experimental paradigm . . . . . . . . . . . 14 2. Means and standard deviations of PPT for all five groups (in minutes) . . . . l9 3. Means and standard deviations of NST for all five groups (in minutes) . . . . 19 4. Means and standard deviations of log NST for all five groups . . . . . . . . . . 20 5. Values of Dunnett's t statistic for individual comparIsons between the‘ control group and each of the ex— perimental groups . . . . . . . . . . . 21 LIST OF FIGURES Figure Page 1. Diagram of the circuit board . . . . . . . . 12 vi LIST OF APPENDICES Appendix Page Appendix A O O O O O O I O O O O O O O O O O O O 3 1 vii INTRODUCTION Duncker (1945) originally proposed that problem solving may be inhibited by the functional fixedness of the solution object. To put it in other words, prior use of an object for its usual use prevents an individual from using that object in an unusual manner in order to solve a problem. To test this proposal he constructed five problems: the gimlet problem, box problem, pliers problem, weight problem, and paper clip problem. For all but the weight problem, he found that preutilization of an object for its normal use made it more difficult for an individual to solve the problem which required an un- usual use of the object as part of its solution. Duncker's work was criticized by Adamson (1952) because of the small number of §§ used and because there was some confounding between groups which were given pre- utilization and those which were not. In addition, Duncker failed to report any statistical tests of significance. Adamson repeated Duncker's experiment with a larger N and more carefully controlled conditions using the box, gimlet, and paper clip problems. His findings were identical to Duncker's, showing that for all three problems those who had prior experience with the critical solution object in its usual manner took longer to solve the problem than those who were not given any preutilization with the cri- tical solution object. Birch and Rabinowitz (1951) in- vestigated the effects of specific experience with objects (a switch or a relay) in unrelated situations upon their utilizability as problem-solving instruments in Maier's two-string problem. They fouhd that §§ who had prior ex- perience with the relay tended to use the switch to solve the problem and that those who had prior experience with the switch tended to use the relay to solve the problem while §§ in a control group (no prior experience with either object) chose each object equally often. They concluded that "what appears to be important for problem solving is not that an individual's performance is depen- dent upon past experience per se but rather that different kinds of experience are differentially effective in influ— encing the content of problem solving behavior." They suggest that two kinds of learning may be important for problem solving: (1) the acquisition by g of certain broad non-specific general notions about the properties of the object or method experienced; and (2) acquisition of experiences which convert the initial perception of broad general properties of an object into perceptions of specific limited functional characteristics. It would seem then that the availability of the critical function of an object plays an important part in whether or not an individual is able to solve a problem. Saugsted (1955), and Saugsted and Raaheim (1957) asked college students to list as many functions as possible for various objects which were later presented with either Maier's candle problem or Saugsted's newspaper problem. They found that those §§ who had given evidence that the necessary functions were available solved the problems almost 100% of the time while those §§ who did not indi— cate that the necessary functions were available solved the problems no more than 58% of the time. They concluded that "a subject will solve any problem if he has the necessary functions . . ." Saugsted and Raaheim (1959) found similar results using sixth-grade children. The suggestion of Birch and Rabinowitz (1951) that the acquisition of certain experiences may convert the initial broad general properties of an object into perceptions of specific limited functional characteris- tics is akin to saying that the availability of functions is decreased by prior or repeated perceptions of specific functions of an object. This is related to Saugsted's finding that those persons having the necessary functions available are less functionally fixed. More support for the availability notion is sup— plied by other studies (Van de Geer, 1957; Duvall, 1965; and Yonge, 1966). All of these studies found that when an unusual usage of an object preceded the presentation of a problem whose solution involved using that object in a normal manner, §§_were not functionally fixed with regard to the usual use of the object. It seems likely that when an unusual use comes first, the availability of broad general properties of an object are retained. In the box problem situation when §§ are asked to recall objects available in the problem situation, solvers report the "box" earlier in their lists than do non-solvers. The solvers also tend to report the box as a separate verbal unit (e.g., "box") whereas non-solvers often report it as "a box of tacks" (Glucksberg, 1964b). Thus it seems that functional fixedness is due to the loss of availability of an object's possible functions. When this availability is restored, functional fixedness is no longer present. The term "availability" however is rather vague and does not help us to know why certain objects or func- tions are available to the individual. A look at both the S-R and the Gestalt theories of problem solving may help us to understand the notion of availability a little more. Maltzman (1955) drawing on Hullian principle and theoretical terms uses the model of a compound habit family hierarchy to explain the problem solving process. He states that the problem elicits a hierarchy of habit families (each habit family itself is composed of a hier- archy of responses). This notion of a compound habit family hierarchy is based upon the assumption "that the elementary laws of behavior derived from conditioning and applicable to trial and error and discrimination learning are also applicable, at least in part, to primary problem solving or reasoning and thinking in general" (Maltzman, 1955). According to the theory an individual goes about solving a problem by responding, first with the most domi- nant response in the most dominant habit family of the compound habit family hierarchy. Changes in the order of dominance in a compound hierarchy may occur due to one or more of the following reasons: (1) the effective reaction potential of the incorrect dominant habit families and their individual members may be decreased as the result of extinction (which would occur when the initial response does not lead to problem solution), (2) an increase in the effective reaction potential of a habit family or families which were initially low in the hierarchy (and which con- tain correct responses) may occur as a result of previous reinforcement of individual members of the hierarchy in that situation. It is important to note that through mediated generalization the decreases or increases in reaction potential just mentioned may generalize to other members of the hierarchy. Thus, in terms of S-R theory, availability may be thought of as referring to the dominance of the correct response. The more dominant the correct response is in an individuals response hierarchy the quicker he should solve the problem. Functional fixedness problems are probably hard for most people to solve because the cor- rect response is probably very low in their response hierarchy in relation to the other possible responses they could make (the other incorrect responses have been repeatedly reinforced either by natural experiences or experimentally induced preutilization). A study by Adamson and Taylor (1954) lends some support to the S-R interpretation of functional fixed- ness. They showed that experimentally induced functional fixedness decreases with time. They suggest that this decrease is due to retroactive interference. Stimulus- reponse theory would explain this decrease as being due to the extinction of the experimentally reinforced habits (or habit families) which first produced functional fixedness. Gestalt explanation of the problem solving process is somewhat different. Gestalt theory puts more emphasis upon the dynamics of the organism than does S-R theory (although present-day mediational S—R theory does not subscribe to machine theory as rigidly as the older non- mediational S-R theory). According to Kahler (1947) the individual does not respond to stimulation, rather he im- poses a pattern of organization upon the stimuli and re- sponds to the products of this organization. A problem occurs when the organization imposed by the individual does not produce the correct response. The problem solving process involves continuous reorganization of the stimuli into new patterns until one of them yields the correct response (solution). How does such organization come about? Gestalt theory states that the individual organizes on the basis of the laws of closure, continuation, simi- larity, proximity, good form, and fittingness (Kohler, 1947; Koffka, 1935). Gestalt theorists do not stress learning to the extent that S-R theorists do. For Gestalt theory learning does not precede original organization but follows it. According to Kohler "association is necessary for recall, and association presuppOses a sufficient de- gree of unification in the sense of organization. Recall, on the other hand, can occur only if the process which is now given resembles some region within the organized trace of the whole experience" (p. 171). For the Gestalt theo- rist functional fixedness results because the function is fixed to a larger pattern (Duncker, 1945) and it is not easy for the individual to reorganize the stimuli into a pattern which produces the correct response. The purpose of this experiment is to see whether or not functional fixedness can be dissipated by means of conceptual grouping tasks, which emphasize either (a) the unusual uses (functions) of various objects or (b) the unusual uses (functions) of various objects related to the critical function in the problem solving task. A conceptual grouping task is one in which an in- dividual is given a number of objects and asked to put all those objects which satisfy a specific criterion into a separate group. It is assumed that when an individual performs such a task he is actively attending to the critical feature. This experiment then, involves the integration of two different research trends (concept formation and problem solving). On the basis of S-R theory one would predict that those §§ who receive conceptual grouping tasks which stress the unusual use of objects for a func- tion which is also a critical function in a problem solving task will solve the problem more quickly than control §§ who do not receive such practice, because the correct response to the problem should become relatively more dominant in the hierarchies of those §§ who have been reinforced for grouping unusual objects together to serve the critical function. Also, decreases in solution time should be a direct function of the amount of prac- tice with such types of grouping. Gestalt theory would predict the same thing. Gestaltists would say that if the way the individual orga- nizes or reorganizes the problem is similar to the organi- zation experienced during the conceptual grouping task then there will be a greater tendancy to recall the pre- vious pattern of organization and use it to organize the stimuli now present. Thus practice on conceptual grouping tasks, which stress the unusual use of objects in a prob- lem solving task, should lead to faster problem solution. Decreases in solution time would likewise be a direct function of the amount of practice with such types of grouping (or organization) because the Gestalt theory states that the accessibility of recall is dependent upon the amount of previous repetition (Koffka, 1935, p. 545). The hypotheses of this experiment are the following: (1) Practice with conceptual grouping tasks which stress only the unusual uses of objects should not make an individual less functionally fixed. Both S—R and Gestalt theory would predict this. S-R theory would say that because such practice is not related to the critical function there would not be enough generalization between such practice and correct response and thus the cor- rect response would remain low in the hierarchy. Gestalt theory would maintain that the similarity (2) 10 between the organization produced by such prac- tice and the organization produced by the problem would not be great enough to aid recall. Practice with conceptual grouping tasks which stress unusual uses of objects for a function which is also the critical function in the prob- lem solving task should reduce the amount of time needed to solve the problem. Furthermore the amount of time needed to solve the problem should be a direct function of the amount of practice on such conceptual grouping tasks. METHOD Subjects: One hundred males who were introductory psychology students at Michigan State University served as §§° None of these §§ had been exposed to the concept of functional fixedness at the time the experiment took place. Experimental Task: Subjects were required to construct a simple electrical circuit (see Figure 1) consisting of two D cell batteries, a SPST switch, and a 6v. bulb. All components were mounted on an 8 in. by 12 in. rectangular masonite pegboard, and the batteries and bulb were pre-wired to four plastic binding posts. The S had to make connections between the binding posts and the screw terminals on the switch. Subjects were given four lengths of wire which combined were 1-1/4 inches too short to complete the de- sired circuit. A 5-1/2 inch screwdriver with a 2-1/2 inch metal blade and a 12-inch coil of plasticene modeling clay, 1/4 inch in diameter were also provided. The prob- lem is solved by using the metal screwdriver blade to compensate for the shortness of wire. This problem is 11 12 .Uumon uflsouflo may no Emummflola.a musmflm 13 known as the circuit problem and was first designed by Glucksberg (1964a). Procedure: The materials for the problem are placed in stan- dard positions on a table and covered with a flexible sheet of Opaque plastic. Posted on a wall to the SL5 left is a simplified diagram of the circuit to be constructed. The §§ are seated at the table and asked to explain the circuit diagram. After indicating that the diagram makes sense to them they are told: "Under this cover is a board with batteries, a bulb, and a switch mounted on it. The problem is to complete the circuit so that the opera— tion of the switch controls the bulb. You may not move any of the components mounted on the board. You may use any of the materials provided with the board." Subjects were divided into five equal groups on the basis of the order of their appearance at the labora- tory. The five groups consisted of a control group (C) and four experimental groups (El, E2, E3, E4). Table 1 shows the paradigm for the present experiment. Before being presented with the circuit problem all five groups were given practice grouping various ob- jects. A11 §§_were asked to make groupings from 10 dif- ferent sets of seven objects. For the experimental groups the groupings stressed unconventional functions. 14 Table l.--Experimental paradigm. Group C Attribute grouping Circuit problem El Iggczlggiiigqugiggtions) Circuit problem E2 Igfigliggglzgglggiggcal functions) Circuit problem E3 €§32i332214gziggiggcal functions) Circuit problem E4 Functional grouping (includes 4 electrical functions Circuit problem one with identical screwdriver) None of the groupings for the E1 group were in any way re- lated to an electrical function. Group E2 received the same grouping as E1 except that the last two groupings presented were directly related to an electrical function (conduction or insulation). Groups E3 and E4 were both similar to group E2 except that their last four groupings were directly related to an electrical function. In addi- tion, a screwdriver—-which was identical to the one pre- sented with the circuit problem--was included in the tenth grouping task for those §§ in group E4. The control group's (C) 10 grouping tasks stressed physical attributes rather than functions. They might be asked for example to group all those objects together which are more than 3 inches long. The grouping tasks of all conditions (C, E1, E2, 15 E3, E4) had been equated for time on the basis of previous pilot work, the average grouping time for each group being nine and one-half minutes. A list of the individual group- ing tasks for each of the five groups may be found in Appen— dix A. Instructions to Group C.--"I am going to show you a number of objects. Some of these objects have something in L.- i ". common. I will tell you which characteristic these objects have in common and you are to take all those objects which . you think have that characteristic and put them together in a group. For example, if I present you with a string, cork, rope, metal bar, a shoelace, a thin piece of wire, and a hammer, I might ask you to put all those objects in a group which weigh less than a pound." Instructions to Groups El' E2, E3, and E4.--"I am going to show you a number of objects. Some of these ob- jects have something in common. I will tell you which function these objects have in common and you are to take all the objects which you think are capable of serving that function and put them together in a group. For example, if I present you with a string, cork, rope, metal bar, a shoelace, a thin piece of wire and a hammer, I might ask you to put all those objects in a group which could be used to tie up a small bundle of magazines. The correct grouping would include the string, rope, shoelace, and the thin piece of wire." 16 If any S failed to include all the correct objects or included an incorrect object, this was pointed out to him and he was asked to restructure the group until he arrived at the correct grouping. The circuit problem was presented immediately after the conceptual grouping tasks. At the beginning of the experiment §§ were told that they were going to ”_r participate in two different experiments. This was done so that 83 would not think there was a connection between __. — x the two parts of the experiment. The conceptual grouping b task was presented as the first experiment and the circuit problem as the second experiment. Measures: The first measure of performance recorded was problem perception time (PPT). Problem perception time was the time taken for S to indicate awareness that in- sufficient wire was available. The time at which any of the following was first observed served as this measure: any verbalization that more wire was needed (in this case E replied that no more wire was available), that more wire would be helpful, any attempt to move components mounted on the circuit board, or any attempts to use modeling clay as a conductor, including questions about the conductivity of clay, or verbalizations such as "So that's it!" or "Can this really be done?" This measure 17 may indicate possible differential effects of the group- ing tasks upon §§ approach to the problem, i.e., does one particular grouping task lead §§ to suspect a tricky problem? A second measure used (a measure of functional fixedness) was the amount of time an S took to solve the circuit problem after he perceived it. This was referred to as net solution time (NST). Subjects were given 20 minutes to solve the problem after they actually perceived it. The problem was considered solved when the bulb was lit. RESULTS One S was discarded because he demonstrated a com- plete lack of comprehension of the problem. Another g was run in his place. Six §§ (two in group E1 and one in each of the other four groups) placed the screwdriver into the circuit without prior indication that they knew they were short of wire. Each of these Sg' total solution time was considered as the best estimate of problem perception time, yielding net solution times of zero for those six §§° This proce- dure follows Glucksberg's (1964a) method of analyzing data for the circuit problem. Three §§ (one in group C and two in group El) failed to solve the problem within the 20 minute time limit. These §§ were assigned NSTs of 20 minutes. Table 2 shows the means and standard deviations for the problem perception times of each of the five groups. A one-way analysis of variance across the five groups was not significant (F < 1, df = 4, 95). Means and standard deviations of net solution times of each of the five groups are presented in Table 3. Be- cause of the heterogeneity of variance as indicated by an 18 19 Table 2.--Means and standard deviations of PPT for all five groups (in minutes). Group E S-D- C 1.61 0.81 El 1.99 1.11 E2 1.78 0.86 E3 1.99 0.92 E4 2.12 1.11 Table 3.--Means and standard deviations of NST for all five groups (in minutes). Group x S.D. C 5.23 5.08 El 5.41 5.38 E2 3.38 2.40 E3 2.74 2.93 B4 2.75 2.76 and Fmax test, (Fmax = 5.02, p < 01) the NSTs were trans- formed into log scores (loglo) as suggested by Edwards (1961). The resulting means and standard deviations of the log NST are shown in Table 4. The Fmax test on the log scores did not show a significant difference in vari- ance (Fmax = 1.53). Therefore the log transformation 20 resulted in approximate homogeneity of variance between groups. Table 4.--Means and standard deviations of log NST for all five groups. Group f S.D. C .5337 .4244 E1 .5605 .3944 1 E2 .4086 .3472 , 4 E3 .2573 .3820 E4 .2649 .4038 A one-way analysis of variance showed a significant difference between the log NST's of the five groups (F = 2.69, df = 4, 95, MSe = .153, p < .036). Dunnett's 3 statistic was applied in making a planned comparison of the NSTs of the experimental groups with the control group. None of the groups differed signif- icantly from the control. However the differences between the control group and groups E3 and E4 approached signifi- cance (p < .10). The results are presented in Table 5. Scheffe's method for testing differences between all possible pairs of groups was applied to see if there were any differences between groups which were not previously compared. There were no significant differences between any of these groups. 21 Table 5.--Values of Dunnett's E statistic for individual comparisons between the control group and each of the experimental groups. Group Dunnett's 5 Significance El .22 n.s. E2 1.01 n.s. E3 2.23 p < .10 E 2.18 p < .10 DISCUSSION Since there were no significant differences be- tween groups in regard to PPT, it may be assumed that the different conceptual grouping tasks did not affect the Sg' initial perception of the problem or their initial mode of attacking the problem. The log transformation of NST scores was successful in eliminating the heterogeneity of variance and in providing relatively normal distribution of these scores. Those §§ who were given practice with unusual func- tions which were unrelated to the electrical function of the circuit problem (Group E1) did not differ significantly in NST from those given practice with object attributes (Group C). This result supports the first hypothesis that practice on a conceptual task that stresses unusual func- tions of objects is not sufficient to dissipate functional fixedness. Those §§ who were given practice with unusual functions, some of which were related to the electrical function of the circuit problem, did better than those §§ who had practice with attributes. Although none of these differences were significant, the differences between 22 23 the control group and those groups that experienced four groupings of objects based on an unusual electrical func- tion (Groups E3 and E4) approached significance (p < .10). It will also be noticed that there is a trend (although not significant) for the NSTs to be inversely proportional to the amount of practice an individual has in grouping on the basis of the critical function. Although the second hypothesis was not confirmed, the results show a trend in the predicted direction. The 3 mean NST of the control group was almost double the mean NST of those groups that received four groupings that were associated with an electrical function. Therefore both the S-R and Gestalt theories mentioned earlier are still very pertinent to this problem. In a problem such as the circuit problem there are many things which may add to the variance of time scores (e.g., mechanical ability, manual dexterity, etc.). The writer feels that if a simpler (in the sense of requiring less complex operations on the part of the individual) problem could be constructed it might be possible to get measurements of solution time which would be much more precise. With such a problem the hypotheses of this study might be confirmed because the variance within groups would be reduced. Stimulus-response theory would explain the trend in reduction of NSTs as due to a strengthening of the low 24 (but correct) response in response hierarchy the low re- sponse becomes strengthened (or becomes more dominant) due to the reinforcement given when the individual correctly groups metal objects as being capable of conducting elec- tricity, or non-metal objects as being incapable of con- ducting electricity. The fact that the screwdriver itself was not presented in the grouping task (except for Group E4) 1 A does not matter since such reinforcement supposedly l generalizes to all objects which are metal or have metal parts. Since the number of reinforcements affects the de— gree to which the response is trengthened, S-R theory would predict that individuals with a greater amount of practice with the critical function would have faster NSTs. We have seen that such a trend is shown by the data. Gestalt theory likewise is capable of accounting for the trends observed. Gestalt theory would predict that the more experience an individual has had with the correct organization of the critical function and the more similar the organization of the critical function is to the individual's organization of the problem, the sooner the person will arrive at the correct reorganization of the problem (which will cause the correct response).. It should be noted that the above theories are not as much at odds as it might seem. The S-R theories could incorporate the organizational processes within their mediational framework and the Gestaltists seem to accept 25 the existance of similarity (generalization) and repetition as playing a part in reorganization and in responding to organization. In reference to the conceptual grouping tasks which were presented to try to make the critical function more available to the individual one thing should be pointed out. It is that this function was made available by giving S5 objects which normally are not used for this function. It is doubtful whether the above grouping tasks would have made this function available if the objects presented were normally used for such functions. If gs had been presented with objects which are normally used to conduct electricity such as switches or relays, it seems improbable that such an experience would have reduced functional fixedness on the circuit problem. The unusual use stresses certain broad general prOperties of the object whereas using an object in its normal manner stresses specific limited functional char- acteristics. It would not be very difficult to test the above notion empirically. It would also be interesting to examine the effect of preutilization on only those individuals who previously have shown evidence of having the necessary functions available. It is reasonable to expect that those who listed the necessary function early in the list would have a higher availability strength and thus would be more re- sistant to functional fixedness (for the particular objects 26 used) than those who listed the necessary functions late in their list (in terms of S-R theory, less dominance should cause less resistance). If this is actually the case, it would parallel evidence by Glucksberg (1964b). Though it might be tempting to attempt to engage in a more theoretical discussion of the processes which play a major part in functional fixedness, the writer thinks that this would be inapprOpriate since the results of this study do not provide any absolute basis on which to base such a discussion. At this point the writer agrees with Duncan's statement that "although theoretical developments are not necessarily unwelcome, the basic need in problem solving is experimental determination of the functional relationships between dimensionalized independent variables and problem solving performance" (1959; p. 425). Ray (1955) and Underwood (1952) have expressed a similar vieWpoint. SUMMARY The purpose of this experiment was to examine the effect of conceptual grouping according to unusual func- tions on functional fixedness, using Glucksberg's circuit problem. Prior to being asked to solve the circuit prob- lem §§_were randomly divided into five equal groups. Each group received practice on a conceptual grouping task. Subjects in the control group (C) were asked to group ob- jects on the basis of a common attribute. All four experi- mental groups (El, E2, E3, E4) were asked to group objects on the basis of unusual functions. The experimental groups differed in the extent to which they were asked to make groupings which stressed an electrical function (this was the critical function of the circuit problem which they were asked to solve after completing the conceptual group- ing task). On the basis of S-R theory and Gestalt theory it was hypothesized that: (1) Practice with conceptual grouping tasks which stress only unusual use of objects (i.e., one which does not stress the critical function) will not reduce functional fixedness. 27 28 (2) Practice with conceptual grouping tasks which stress unusual use of objects for the critical function should reduce the functional fixedness in the circuit problem. Furthermore the amount of time neededlto solve the problem should be a direct function of the amount of practice with the critical function. The results supported the first hypothesis. The 1 second hypothesis was not confirmed but there was a strong trend in the predicted direction. The results were con- E sistant with either S-R or Gestalt theory. REFERENCES Adamson, R. E. Functional fixedness as related to problem solving: a repetition of three experiments. Jnl. Adamson, R. E. and Taylor, D. W. Functional fixedness as related to elapsed time and to set. Jnl. Exp; Psychol., 1954, 41, 122-126. Birch, H. G. and Rabinowitz, H. S. The negative effect of previous experience on productive thinking. Jnl. Exp. Psychol., 1951, 41, 121-125. Duncker, K. On problem solving. Psychol. Monogr., 1945, 58, no. 5 (whole no. 270). Duvall, A. N. Functional fixedness: a replication study. Psychol. Rec., 1965, 15, 497-499. Edwards, A. L. Experimental Design in Psychological Re- search. Rev. ed., Holt, Rinehart and Winston, New York, 1965. Glucksberg, 8. Problem solving: response competition and the influence of drive. Psychol. Rep., 1964a, 15, 939-942. Glucksberg, S. Effects of verbal behavior on problem solving: labeling the functionally fixed object. Amer. Psychologist, 1964b, 19, 575 (abstract). Koffka, K. Principles of Gestalt Psychology. Harcourt, Brace and Co., New York, 1935. K6hler, W. Gestalt Psychology, Liveright Publishing Corp., New Yofk, I94 . Maltzman, I. Thinking: From a behavioristic point of View. Psychological Rev., 1955, £2, 275-286. Ray, W. S. Complex tasks for use in human problem-solving research. Psychol. Bull., 1955, 52, 134-149. 29 30 Saugsted, P. Problem-solving as dependent on availability of functions. Brit. J. Psychol., 1955, 35, 191-198. Saugsted, P. and Raaheim, K. Problem-solving and avail- ability of functions. Acta Psychol., 1957, $5, 263- 278 Saugsted, P. and Raaheim, K. Problem-solving and avail- ability of functions in children. Acta Psychol., 1959, 55, 45-58. Underwood, B. J. An orientation for research on thinking. Psychol. Rev., 1952, 55, 209-220. Van de Geer, J. P. A psychologigal study of problem-solving. Haarlem: Uitgeverij De Toorts, 1957. Yonge, G. D. Structure of experience and functional fixed- ness. Jnl. Educ. Psychol., 1966, 51, 115-120. APPENDIX A Grouping tasks for Group C (the correct objects are underlined). Group together all those objects which: 1. are flexible crayon, rOpe, feather, wrench, ashtray, wire, paper clip. weigh less than one ounce £222! ruler, 3255, book, pocket knife, £123! pgp. are more than three inches long window handle, coal, nut, nail, Clothespin, crayon, cork. are very brittle spoon, cork, gagik, screw, coke bottle, shoe- lace, book. weight less than one-half ounce pin, nut, compass, cork, paper clip, pocket- knife, nail. are less than one inch long nut, button, coal, small paper clip, large paper clip, feather, quarter 31 used: 10. Grouping tasks for Group E 32 are combustable quarter, rope, coal, door knob, wooden ruler, book, screw weigh more than three ounces, book, fork, pliers, door knob, window handle, can opener, wrench. are less than seven inches long fork, pliers, spoon, compass, coke bottle, pen, book. are circular in one way or another cork, hammer, coke bottle, pgncil, coal, round ashtray, nail. 1 Group together all those objects which can be 1. to lace up your shoes string, cork, twine, metal bar, shoelace, E912 yipg, hammer. to write with crayon, chapstick, fork, chalk, pliers, pencil, coal. to hold a small sheaf of papers together crescent wrench, clip folder, pencil, squeeze Clothespin, paper clip, spoon, quarter. 10. 33 to help draw a circle compass, feather, thumbtack, string, ring, hammer (round head), coke bottle. to pound a fairly small nail into a board coke bottle, heavy glass ashtray, hammer, crescent wrench, ruler, door knob, metal bar. to hang your winter coat 9p pencil, coathanger, thumbtack, paper clip, nail, door knob, ashtray. as an inkwell cork (partly hollow), ashtray, fork, quarter, door knob, ring, spoon. as a bookmark feather, walnut, string, nail, coal, can opener, paper clip. to pin a notice on an ordinary cork bulletin board pocket knife, nail, spoon, paper clip, fork, pliers, compass. to stir a cup of coffee paper clip, table knife, door knob, window handle, string, pliers, wax crayon. 34 Grouping tasks for Group E2- Group together all those objects which can be used: 1. to lace up your shoes string, cork, twine, metal bar, shoelace, £912 wire, hammer. 2. to write with crayon, chapstick, fork, chalk, pliers, pencil, h coal. 3. to hold a small sheaf of papers together crescent wrench, clip folder, pencil, squeeze Clothespin, paper clip, spoon, quarter. 4. to help draw a circle compass, feather, thumbtack, string, ring, hammer (round head), coke bottle. 5. to pound a fairly small nail into a board coke bottle, heavy glass ashtray, hammer, crescent wrench, ruler, door knob, metal bar. 6. to hang your winter coat op pencil, coathanger, thumbtack, paper clip, nail, door knob, ashtray. 7. as an inkwell cork (partly hollow), ashtray, fork, quarter, door knob, ring, spoon. 8. as a bookmark feather, walnut, string, nail, coal, can opener, paper cliJ . 9. 10. 35 to conduct electricity wire, glass ashtray, fork, pliers, pencil, paper clip, nail. to insulate against electricity rubber cork, glass ashtray, wrench, spoon, crayon, metal bar, thumb tack. Grouping tasks for Group E3. Group together all those objects which can be used: 1. to lace up your shoes string, cork, twine, metal bar, shoelace, thin wire, hammer. to write with crayon, chapstick, fork, chalk, pliers, pencil, coal. to hold a small sheaf of papers together crescent wrench, clip folder, pencil, sgueeze Clothespin, paper clip, spoon, quarter. to help draw a circle compass, feather, thumbtack, string, ring, hammer (round head), coke bottle. to pound a fairly small nail into a board coke bottle, heavy_glass ashtray, hammer, crescent wrench, ruler, door knob, metal bar. 10. 36 to hang your winter coat op pencil, coathanger, thumbtack, paper clip, nail, door knob, ashtray. to conduct electricity wire, glass ashtray, fork, pliers, pencil, paper clip, nail. to insulate against electricity rubber cork, glass ashtray, wrench, spoon, crayon, metal bar, thumb tack. both to insulate against and also conduct electricity (i.e. the object must be capable of doing both of these thing, not only one). hammer, coal, spring Clothespin, string, Chalk, metal clip folder, pocket knife with plastic handle. to conduct electricity coke bottle, compass, button, window handle, table knife, shoelace, coathanger. Grouping tasks for Group E4. Group together all those objects which can be used: 1. to lace up your shoes string, cork, twine, metal bar, shoelace, thin wire, hammer. 37 to write with crayon, chapstick, fork, chalk, pliers, pencil, coal. to hold a small sheaf of papers together crescent wrench, Clip folder, pencil, sgueeze Clothespin, paper Clip, spoon, quarter. to help draw a circle compass, feather, thumbtack, string, ring, hammer (round head), coke bottle. to pound a fairly small nail into a board coke bottle, heavy glass ashtray, hammer, crescent wrench, ruler, door knob, metal bar. to hang your winter coat op pencil, coathanger, thumbtack, paper clip, nail, door knob, ashtray. to conduct electricity .wire, glass ashtray, fork, pliers, pencil, paper clip, nail. to insulate against electricity rubber cork, glass ashtray, wrench, spoon, crayon, metal bar, thumb tack. both to insulate against and also conduct electricity (i.e., the object must be capable or doing both of these thing, not only one) hammer, coal, spring Clothespin, string, chalk, metal clip folder, pocket knife with plastic handle. 38 10. to conduct electricity coke bottle, compass, button, screwdriver,* tableknife, shoelace, coathanger. *The screwdriver is identical to the one provided with the circuit problem. ”)fiififiiJHHIIflfljfli(fljijiflflliflmflmfilfillms