This article was downloaded by: [The University of Texas at Dallas] On: 22 October 2014, At: 11:59 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
The Journal of General Psychology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/vgen20
Negative and Positive Incentive Contrast Effects with Saccharine versus Sucrose Lawrence Weinstein
a
a
University of Ife , Nigeria Published online: 06 Jul 2010.
To cite this article: Lawrence Weinstein (1978) Negative and Positive Incentive Contrast Effects with Saccharine versus Sucrose, The Journal of General Psychology, 98:2, 225-240, DOI: 10.1080/00221309.1978.9920876 To link to this article: http://dx.doi.org/10.1080/00221309.1978.9920876
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
The Journal of General Psychology, 1978, 98, 225-240.
NEGATIVE AND POSITIVE INCENTIVE CONTRAST EFFECTS WITH SACCHARINE VERSUS SUCROSE*',' University of Ife, Nigeria
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
LAWRENCE WEINSTEIN
SUMMARY Forty male white rats received 19 sessions of bar press training with sucrose (16% or 4%) as reward followed by a shift from 4% to 16% and from 16% to 4%. Three dependent measures indicated positive and negative contrast effects. In Experiment II, 70 male white rats received 14 sessions of bar press training with ( 1 . 2 % , 1%, . l o % ,or .01%)saccharine, and water followed by a shiftfrom 1.2% to . l o % .1 . 2 % to . O l % , 1 to . l % ,and .lo%to .01%.Threeof the four measures indicated negative contrast effects. In Experiment 111, 40 male white rats experienced 16 sessions of bar press training with 196, . l o % , .05%, or .01%of saccharine followed by an increase from. 1%. to 196, .05% to 1%, and .01% to 1%. Three of the measures indicated positive contrast effects. Confounding inherent in the use of solid food or sucrose did not appear to account for negative and positive incentive contrast effects. A.
INTRODUCTION
When a decrease in incentive size results in a decrement in performance significantly below a baseline defined by Ss trained only a t the postshift magnitude, the result is referred to as negative incentive contrast effects; when upward shifts in reinforcement magnitude produce performance levels higher than exposure to the larger reward amount alone, positive incentive contrast effects are said to occur. Studies typically use solid food or sucrose solution as incentive with infrahumanSs. A change in amount of solid food or amount of sucrose is quite possibly confounded with a difference in body weight. Few studies have examined the effects of magnitude of reward on body weight (e.g., 5 ) .
* Received in the Editorial Office, Provincetown, Massachusetts, on January 3 1 , 1977. Copyright, 1978, by The Journal Press. I This research was presented at the 1970 Australian Psychological Society meeting in Hobart, August. * Requests for reprintq should be sent to the author at the address shown nt the end of this article. 225
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
226
JOURNALOF GENERAL PSYCHOLOGY
If training under different magnitudes of reinforcement results in weight differences, then all earlier studies which have examined the effects of a change in incentive size on animal behavior would appear to be confounded by the difference in body weight. Experiment I examined changes in incentive magnitude under sucrose, while body weight was monitored. If a decrease in amount of solid food or concentration of sucrose is confounded with a difference in body weight, the source of confounding may be eliminated by using saccharine solutions, which lead to equivalent body weights, as incentives. Collier and Novel1 (2) have demonstrated that consumption of a .01, .03, .30, .90, or 2 . 7 % saccharine solution does not alter body weight compared to a .lo% saccharine solution. Hulse (8)and Vogel, Mikulka, and Spear ( 11) decreased the concentration of a constant volume saccharine solution and failed to obtain negative contrast effects. This may be because they selected preshift concentrations not conducive to the production of negative contrast effects. Many studies have demonstrated that the magnitude of or probability of obtaining negative contrast is a positive function of the amount of reward reduction (e.g., 4 , 6, 12). Hulse (8) decreased the concentration of a saccharine solution from 1% to .1% in an operant conditioning chamber. Vogel et al. reduced the concentration of a saccharine solution from .lo% to .01% following training in an operant conditioning box. Quite possibly, larger reductions in incentive magnitude would produce negative contrast effects. If downward contrast is not obtained with an appropriate decrease in the concentration of a saccharine solution when body weights of the saccharine groups are statistically the same, then all previous demonstrations of negative incentive contrast effects with infrahuman organisms would appear to be due to a confounding inherent in the use of solid food or sucrose. Experiment I1 examined a reduction in reward magnitude under saccharine. If increments in reward quantity are confounded with a difference in body weight, the confounded variable might be removed by using saccharine solutions as incentives. Hulse (8) increased the concentration of constant volume saccharine solutions from .05 and .1% to 1 .O% and failed to obtain positive incentive contrast effects. As already discussed, this may be due to his selection of preshift concentrations. If positive contrast effects are not obtained with a suitable increase in the concentration of a saccharine solution when body weights in the saccharine groups are not reliably different, then it would seem that all previous demonstrations of positive incentive contrast effects using animal Ss are the result of a confounding inherent in the use of solid food or sucrose. Experiment 111 examined the effects of increases in reward magnitude under saccharine.
LAWRENCE WEINSTEIN
227
B. EXPERIMENTI
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
1. Method
a. Subjects. The Ss were 40 experimentally naive, male white rats, 60-90 days old at the start of the experiment and were housed in individual cages. b. Apparatus. The operant conditioning box had plywood side walls, Plexiglass top, back and door, a grid floor, and measured 305.8 x 254 x 254 mm. A rectangular hole 12.7 mm high by 50.8 mm long was cut in the side of the box 127 mm above the floor so that the bar could be inserted 12.7 mrn into the cage. A circular hole 9.4 mm in diameter was cut in the floor. Through this openings had access to the .4 cc cup of the dipper, whenever it was presented. The liquid used was presented to S by a dipper type magazine driven by a motor. The metal dipper rested in a plastic tray which contained the liquid reward; the tray was mounted below the hole in the floor, underneath the box, so that whenever S pressed the bar, the cup of the dipper travelled 50.8 mm upwards and was presented for 2.0 sec during which bar presses were recorded and did not operate the dipper. Two dim house lights illuminated the box. Events were programmed and recorded automatically by timers, counters, etc. located in the experimental room. c. Procedure. Levels of sucrose were selected from the literature. Guttman (7) reported maximum and minimum bar pressing to a 16% and 4% sucrose solution, respectively; furthermore, Weinstein ( 13)demonstrated contrast effects with 4 and 16% sucrose solutions. Ten Ss were randomly assigned to each of four equal groups. Sucrose concentration treatments of 4% and 16% by weight3 were used with two control groups and two shift groups. The 4% control group (4-4) and the 16% control group (16-16) were run throughout the experiment without changing concentrations. The 4-16 group was shifted on the 20th session from 4 to 16% sucrose, and the 16-4 Ss were changed on the 20th session from 16% sucrose to 4% sucrose. Upon receipt in the laboratory all Ss were given seven days of continuous access to laboratory blocks and water, while weights were recorded daily. Following 15 days of adaptation to a food and water deprivation schedule, Ss were trained on successive days to press the bar for continuous reinforcement. Each session lasted for 30 reinforcements dipper presentations. This regime was continued until performance appeared to be stable, requiring 19 sessions. Concentrations were then changed or not, and all Ss were continued through For example, the 16% gm. of cane sugar with 84 ml. of tap water. The saccharine solutions were prepared in a similar manner.
228
JOURNAL OF GENERAL PSYCHOLOGY
the 25th session. Throughout the study body weights were recorded before each session, and S s had access to dry pellets and water for 1 hr following the experimental period.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
2.
Results
The mean body weight was one measure examined in the analysis of the results. From Figure 1 it seems that initially, for the seven days previous to adaptation, the mean body weight did not differ among the groups. The average body weight over these seven days did not differ reliably among the groups by analysis of variance (F = 2.13, df = 3/36, p > .05). From Figure 2 it appears that for the 15 days of adaptation, the mean body weight did not differ between the groups. By analysis of variance the average body weight from days 1-15 was not statistically different among the groups (F = 1.83, df = 3/36, p > .05). From Figure 3 it seems clear that from session 1-20 the 4-4 Ss weighed the same as the 4-16 group, the 16-16 Ss weighed the same as the 16-4 animals, and the 4-4 and 4-16 groups each weighed less than the 16-16 or 16-4 animals. The mean body weight from sessions 1-20 differed significantly among the four groups by analysis of variance (F = 5.16, df = 3/36, p < .01). By Duncan’s multiple comparisons the differences between 4-4 and 4-16, and between 16-16 and 16-4 were each not statistically significant (p > .05), while the difference between each shift group (4-16 and 16-4) and the respective control groups (16-16 and 4-4) was highly reliable (p < ,005). Four other measures were examined in the analysis of the results: average number of bar presses made during each session, average volume (ml) of liquid consumed during each session, average latency of the fifth bar press per session, average bar presses per min. Since average bar presses per min yielded essentially the same results as the above measures, excluding mean bar press latency, only average bar presses per min are reported here. Figure 4 is illustrative of three measures. From Figure 4 it is apparent that from sessions 1-19, 4% sucrose produced lower rates of bar pressing than 16% sucrose. The mean number of bar presses per min from sessions 1-19 differed significantly among the four groups by analysis of variance (F = 4.62, df = 3/36, p < .01). By Duncan’s multiple comparisons the differences between 4-4 and 4-16, and between 16-16 and 16-4 were each not statistically reliable (p > .05), while the difference between each shift group (4-16, and 16-4) and the respective controls was highly significant (p < .01).
LAWRENCE WEINSTEIN
.---.
.--+
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
O--U
4-4 16-4 16-16 4-16
DAYS MEAN BODYWEIGHTFOR IN
FIGURE 1 THE DAYSPREVIOUS TO ADAETATION, EXPERIMENT 1
229
230
JOURNAL OF GENERAL PSYCHOLOGY
24C
-E
0--0 4-4 M16-4 A-A 16-1 6 A-----A4-16
0,
W
I-
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
I
9 22c W 3
&
0 m Z
WQ 2oc I C
1
1
1
1
l
1
1
1
1
1
1
1
10
5
ADAPTATION DAYS
1
1
1
15
FIGURE 2 MEANBODYW E i C m FOR THE ADAPTATIONDAYS, IN EXPERIMENT I
Figure 4 indicates that on session 20 the shift groups abruptly changed their bar press rates to a level below their respective control groups (negative contrast effects). The mean number of bar presses per min during sessions 20-23 differed significantly among the two shift and two control groups by analysis of variance ( F = 4.62, df = 3/36,p < .01). By Duncan’s comparisons the differences between the shift groups and their respective control groups were each significant (p < .05). All measures except latency showed these significant concentration differences in the preshift and incentive contrast effects in the postshift. C.
EXPERIMENT I1
Experiment I indicated that a shift in incentive magnitude is confounded with a difference in body weight. In Experiment I1 saccharine solutions were
LAWRENCE WEINSTEIN
23 1
2404-4 16-4 =----I 16-16 o----o 4-16 +-a
230Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
lh
E rn W
220 -
=
"U
'0,
I
a-d
190
T 5 20
0 0
1 0
15
SESSIONS FIGURE 3 MEANBODYWEIGHTFOR THE EXPERIMENTAL SESSIONS WITH SUCROSE, I N EXPERIMENT I
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
232
JOURNAL O F GENERAL PSYCHOLOGY
0
5
10
15
20
25
SESSIONS used as incentives to examine the influence of a decrease in amount of reinforcement in the absence of :I confounded variable. 1.
Method
T h e S s were i 0 experimentally naive, male white rats. T h e details of age, housing, a n d apparatus were the same as in Experiment I.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
LAWRENCE WEINSTEIN
233
Levels of saccharine were selected from the studies previously cited (8, 11). O n the basis of a pilot study, it was determined that a 1 . 2 % saccharine solution produces a higher rate of bar pressing than a 1%, .lo%,and .01% saccharine solution. Ten Ss were randomly assigned to each of seven equal groups. Saccharine concentration treatmentsof 1 . 2 % , 1.0%,.lo%, and .Ol%,by weight, and tap water were used with control groups and four shift groups. The water (W-W), .lo% control group ( . l - . l ) , and .01% control group (.01-.01) were run throughout the experiment without changing concentrations. On the 15th session, the 1.2-.10 group was changed from 1 . 2 % saccharine to .lo%, the 1.2-.Ol Ss were shifted from 1 . 2 to .01%,the 1 - . 1 group was decreased from 1.0%to .lo%, and, finally, the .1-.01 group was shifted from . l o %to .01% saccharine. The details of adaptation, bar press training, and deprivation schedule were the same as in Experiment I except that performance appeared stable after 14 sessions. 2.
Results
The mean body weight was one measure examined in Experiment 11. Initially, for the seven days previous to adaptation, and for the 15 days of adaptation there was no statistically significant difference in body weight between the groups. From Figure 5 it seems clear that from sessions 1-15 all groups weighed the same. The mean body weight from sessions 1-15 did not differ significantly among the six groups by an analysis of variance (F = 2.10, df = 5/54,p> .05). The same four measures of Experiment I were examined in the present Experiment. Mean bar presses per min yielded essentially the same results as the other measures except average bar press latency. Only average bar presses per min are reported here. From Figure 6 it seems that from sessions 1-14 the 1 . 2 shift groups had the same rate of bar pressing, and the .10 shift and control groups reached the same level of behavior. Figure 6 further suggests that 1 . 2 % produced a higher rate of bar pressing than 1%, .lo%,and .Ol%;and 1.0%resulted in a higher level of behavior than . l o % saccharine which produced more bar pressing than . O l % saccharine. Furthermore, .01% appears to result in more bar presses per min than water. The mean number of bar presses per min from sessions 1-14 differed significantly among the seven groups by an analysis of variance (F = 2.34, df = 6/63, p < .05). By Duncan’s comparisons the differences between 1 . 2 - . I and 1.2-.01, and .l-.01 and .1-.1 were each not reliable (p > .05), while the differences between W-W and each other group,
JOURNALOF GENERAL PSYCHOLOGY
0--0 1*2- 0.1 .--. 1.2-001 &----A 1.0-0-1
A-----A04 - 001
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
0 - 4 04 - 0-1 [1--0 0.01-001
0
1
0
~
,
,
5
t
1
l
10
l
~
15
SESSIONS FIGURE 5 MEAN BODYWEIGHTI N GRAMSFOR THE EXPERIMENTAL SESSIONS WITH SACCHARINE, I N EXPERIMENT I1
1.2-.1 and 1-.1, andtheshiftgroups(1.2-.1,1.2-.01, 1-.I, and .l-.Ol)andthe respective controls were each reliable (p < .05). Figure 6 indicates that on session 15, groups 1.2-.1 and 1.2-.01 each suddenly decreased its bar press rates to a level below its respective control group (negative contrast effect), while the remaining shift groups reduced their bar press rates to the level of their respective control groups. The mean number of bar presses per min during sessions 15-20 differed significantly among the four shift and two control groups by analysis of variance ( F = 3 . 1 3 , df = 5/54, p < .05). By Duncan’s comparisons the difference between 1.2-.1 and . l - . I , and 1.2-.01 and .01-.01 were each significant (p < .05), while the differences between 1-.I and .1-.1, and .l-.01 and .01-.01 were each not statistically reliable (p > .OS).
l
1
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
LAWRENCE WEINSTEIN
5 MEANBARPRESSES
235
-
15
10 SESSIONS PER
FIGURE 6 MINUTEFOR SACCHARINE,
IN
EXPERIMENT 11
20
23 6
JOLRNAI. OF GENERAL PSYCHOI,O(;Y
All other mea3ures except average bar press latency indicated these significant concentration differences in the preshift and negative contrast effects in the postshift.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
D.
ESPERIMENT111
Experiment I1 indicated that negative contrast effects can be obtained with appropriate concentrations of saccharine solution which do not lead to weight differences. In Experiment 111 saccharine solutions were again used as reinforcement .05), while the difference between .01-1 and 1-1 was reliable Cp < .05).
E.
DISCUSSION
Experiment I clearly demonstrated that a change in incentive magnitude which results in negative and positive incentive contrast effects is confounded with a difference in body weight.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
LAWRENCE WEINSTEIN
23 9
The demonstration of negative contrast effects with a reduction in the concentration of a sucrose solution is not in accord with a number of studies in the literature (e.g., 5 , 7 , 10). The discrepant findings may be due to the different deprivation conditions selected by the present investigation and the three studies previously cited. In the present study, Ss were deprived of both food and water, while Dunham and Kilps, Guttman, and Rosen deprived their S s of either food or water. Quite possibly the present more severe deprivation schedule is more likely to produce negative contrast effects than food or water deprivation alone. In Experiment I the finding of positive contrast with an increase in incentive magnitude disagrees with many stuhes in the literature (e.g., 1,3,9). The discrepant findings, are, perhaps, accounted for by the more severe deprivation conditions employed in the present study as compared to the Ashida et al., Collier et al., and Metzger e t al. investigations. Experiment I1 clearly demonstrated negative contrast effects with a reduction in concentration of a constant volume saccharine solution. This result disagrees with Hulse (8) and Vogel et al. (11). As already discussed, the disparate findings are most likely due to the different preshift concentrations selected by the two studies previously cited and the present investigation. The finding that the water group responded a t a significantly lower level than all other groups indicates that the concentrations used were all reinforcers; none was a punishing agent (where a reinforcement is defined as an event which produces a significantly higher level of performance as compared to a control g r o u p i n this case the Ss exposed to water-which does not experience the event). Since negative contrast effects were not obtained with the saccharine concentrations employed by Hulse ( 1 .O and .lo’%) and Vogel et 01. (. 10 and .01%), the present contrast effects are quite probably due to the particular concentrations chosen, 1.2% and . O l % , and cannot be accounted for by differences in the present test situation and those situations used by Hulse or Vogel et al. Since negative contrast effects are obtained with saccharine solutions which do not lead to different body weights, the phenomenon cannot be explained on the basis of a confounding inherent in previous research with animals. The results of Experiment I11 indicate that positive incentive contrast effects are obtained with an increase in concentration of a saccharine solution. This result disagrees with Hulse (8).As already discussed, the divergent findings are very probably not due to different test situations, since, in the present study, contrast was not obtained with the concentrations used by Hulse; the
240
JOURNALOF GENERAL PSYCHOLOGY
contrasting findings are more probably due to the different preshift concentrations selected by Hulse and the current investigation. Since positive contrast effects are obtained with an increment in incentive magnitude in the relative absence of confounding factors, the phenomenon cannot be accounted for by adifference in body weight inherent in earlier animal research,
REFERENCES 1.
ASHIDA,S.. & BIRCH,D. The effects of incentive shift as a function of training. Psychor;.
2.
COLLIER,C. H . , & NOVELL,K . Saccharine as a sugar surrogate. 1.Comp. 6. Physio.!. Psyrhol., 1967, 64, 404-408. COLLIER,G . , KNARR,F. A , , & MARX. M. H . Some relations between the intensibe properties of the consummatory response and reinforcement. 1.Exprr. P s y r h o l . , 196 I . 62, 484-495. DILOLLO,V. D., & BEEZ,V . Negative contrast effect as function of magnitude of reward decrement. Psyrhon. S r i . , 1966, 5, 99-100. DUNHAM,P. J . , & KILPS,B . Shifk in magnitude of reinforcement: Confounded factors or contrast effects? J . Exper. Psychol., 1969, 79, 373-374. GONZALEZ,K. G., GLEITMEN, H . , & BITTERMAN, M. E. Some observations on the depression effect. J. Comp. Er Physiol. Psychol., 1962, 55, 578-581. GUTTMAN,N . Operant conditioning, extinction, and periodic reinforcement in relation to concentrationofsucroseusedasreinforcingagent.J. Exper. Psyrhol., 1953,46,2 13-222. HULSE,S. H. Partial reinforcement, continuous reinforcement, and reinforcement shift effects. J. Exprr. Psyrhol., 1962, 64, 451-459. METZCER,R . , COTTON,J. W . . & LEWIS,D. J . Effect of reinforcement magnitude and order of presentation of different magnitudes on runway behavior. J . Comp. Er Physiol.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
S c i . , 1964, 1, 201-202.
3.
4.
5. 6.
7.
8. 9.
Psyrhol., 1957. 50, 184-188.
ROSEN.A. J. Incentive shift performance as a function of magnitude and number of sucrose rewards. J. Comp. Ci Physiol. Psyrhol., 1966, 62, 487-490. 11. VOGEL.J. K. MIKULKA, P. J. & SPEAR,N. E . Effects of shifh in sucrose and saccharine concentrations on ticking behavior in the rat. 1.Comp. Ci Physiol. Psyrhol.. 1968. 66, 10.
~
661-666.
WEINSTEIN. L. Magnitude of incentive contrast w a function of amount of verbal reward change. Psyrhon. S r i . , 1970, 21, 65-66, 13. . Negative incentive contrast with sucrose. Psychon. S c i . , 1970, 19, 13-14.
12.
~
clo John Bahchuk Box 1289 Vegreville, Alberta, Canada
The Journal of General Psychology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/vgen20
Negative and Positive Incentive Contrast Effects with Saccharine versus Sucrose Lawrence Weinstein
a
a
University of Ife , Nigeria Published online: 06 Jul 2010.
To cite this article: Lawrence Weinstein (1978) Negative and Positive Incentive Contrast Effects with Saccharine versus Sucrose, The Journal of General Psychology, 98:2, 225-240, DOI: 10.1080/00221309.1978.9920876 To link to this article: http://dx.doi.org/10.1080/00221309.1978.9920876
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
The Journal of General Psychology, 1978, 98, 225-240.
NEGATIVE AND POSITIVE INCENTIVE CONTRAST EFFECTS WITH SACCHARINE VERSUS SUCROSE*',' University of Ife, Nigeria
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
LAWRENCE WEINSTEIN
SUMMARY Forty male white rats received 19 sessions of bar press training with sucrose (16% or 4%) as reward followed by a shift from 4% to 16% and from 16% to 4%. Three dependent measures indicated positive and negative contrast effects. In Experiment II, 70 male white rats received 14 sessions of bar press training with ( 1 . 2 % , 1%, . l o % ,or .01%)saccharine, and water followed by a shiftfrom 1.2% to . l o % .1 . 2 % to . O l % , 1 to . l % ,and .lo%to .01%.Threeof the four measures indicated negative contrast effects. In Experiment 111, 40 male white rats experienced 16 sessions of bar press training with 196, . l o % , .05%, or .01%of saccharine followed by an increase from. 1%. to 196, .05% to 1%, and .01% to 1%. Three of the measures indicated positive contrast effects. Confounding inherent in the use of solid food or sucrose did not appear to account for negative and positive incentive contrast effects. A.
INTRODUCTION
When a decrease in incentive size results in a decrement in performance significantly below a baseline defined by Ss trained only a t the postshift magnitude, the result is referred to as negative incentive contrast effects; when upward shifts in reinforcement magnitude produce performance levels higher than exposure to the larger reward amount alone, positive incentive contrast effects are said to occur. Studies typically use solid food or sucrose solution as incentive with infrahumanSs. A change in amount of solid food or amount of sucrose is quite possibly confounded with a difference in body weight. Few studies have examined the effects of magnitude of reward on body weight (e.g., 5 ) .
* Received in the Editorial Office, Provincetown, Massachusetts, on January 3 1 , 1977. Copyright, 1978, by The Journal Press. I This research was presented at the 1970 Australian Psychological Society meeting in Hobart, August. * Requests for reprintq should be sent to the author at the address shown nt the end of this article. 225
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
226
JOURNALOF GENERAL PSYCHOLOGY
If training under different magnitudes of reinforcement results in weight differences, then all earlier studies which have examined the effects of a change in incentive size on animal behavior would appear to be confounded by the difference in body weight. Experiment I examined changes in incentive magnitude under sucrose, while body weight was monitored. If a decrease in amount of solid food or concentration of sucrose is confounded with a difference in body weight, the source of confounding may be eliminated by using saccharine solutions, which lead to equivalent body weights, as incentives. Collier and Novel1 (2) have demonstrated that consumption of a .01, .03, .30, .90, or 2 . 7 % saccharine solution does not alter body weight compared to a .lo% saccharine solution. Hulse (8)and Vogel, Mikulka, and Spear ( 11) decreased the concentration of a constant volume saccharine solution and failed to obtain negative contrast effects. This may be because they selected preshift concentrations not conducive to the production of negative contrast effects. Many studies have demonstrated that the magnitude of or probability of obtaining negative contrast is a positive function of the amount of reward reduction (e.g., 4 , 6, 12). Hulse (8) decreased the concentration of a saccharine solution from 1% to .1% in an operant conditioning chamber. Vogel et al. reduced the concentration of a saccharine solution from .lo% to .01% following training in an operant conditioning box. Quite possibly, larger reductions in incentive magnitude would produce negative contrast effects. If downward contrast is not obtained with an appropriate decrease in the concentration of a saccharine solution when body weights of the saccharine groups are statistically the same, then all previous demonstrations of negative incentive contrast effects with infrahuman organisms would appear to be due to a confounding inherent in the use of solid food or sucrose. Experiment I1 examined a reduction in reward magnitude under saccharine. If increments in reward quantity are confounded with a difference in body weight, the confounded variable might be removed by using saccharine solutions as incentives. Hulse (8) increased the concentration of constant volume saccharine solutions from .05 and .1% to 1 .O% and failed to obtain positive incentive contrast effects. As already discussed, this may be due to his selection of preshift concentrations. If positive contrast effects are not obtained with a suitable increase in the concentration of a saccharine solution when body weights in the saccharine groups are not reliably different, then it would seem that all previous demonstrations of positive incentive contrast effects using animal Ss are the result of a confounding inherent in the use of solid food or sucrose. Experiment 111 examined the effects of increases in reward magnitude under saccharine.
LAWRENCE WEINSTEIN
227
B. EXPERIMENTI
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
1. Method
a. Subjects. The Ss were 40 experimentally naive, male white rats, 60-90 days old at the start of the experiment and were housed in individual cages. b. Apparatus. The operant conditioning box had plywood side walls, Plexiglass top, back and door, a grid floor, and measured 305.8 x 254 x 254 mm. A rectangular hole 12.7 mm high by 50.8 mm long was cut in the side of the box 127 mm above the floor so that the bar could be inserted 12.7 mrn into the cage. A circular hole 9.4 mm in diameter was cut in the floor. Through this openings had access to the .4 cc cup of the dipper, whenever it was presented. The liquid used was presented to S by a dipper type magazine driven by a motor. The metal dipper rested in a plastic tray which contained the liquid reward; the tray was mounted below the hole in the floor, underneath the box, so that whenever S pressed the bar, the cup of the dipper travelled 50.8 mm upwards and was presented for 2.0 sec during which bar presses were recorded and did not operate the dipper. Two dim house lights illuminated the box. Events were programmed and recorded automatically by timers, counters, etc. located in the experimental room. c. Procedure. Levels of sucrose were selected from the literature. Guttman (7) reported maximum and minimum bar pressing to a 16% and 4% sucrose solution, respectively; furthermore, Weinstein ( 13)demonstrated contrast effects with 4 and 16% sucrose solutions. Ten Ss were randomly assigned to each of four equal groups. Sucrose concentration treatments of 4% and 16% by weight3 were used with two control groups and two shift groups. The 4% control group (4-4) and the 16% control group (16-16) were run throughout the experiment without changing concentrations. The 4-16 group was shifted on the 20th session from 4 to 16% sucrose, and the 16-4 Ss were changed on the 20th session from 16% sucrose to 4% sucrose. Upon receipt in the laboratory all Ss were given seven days of continuous access to laboratory blocks and water, while weights were recorded daily. Following 15 days of adaptation to a food and water deprivation schedule, Ss were trained on successive days to press the bar for continuous reinforcement. Each session lasted for 30 reinforcements dipper presentations. This regime was continued until performance appeared to be stable, requiring 19 sessions. Concentrations were then changed or not, and all Ss were continued through For example, the 16% gm. of cane sugar with 84 ml. of tap water. The saccharine solutions were prepared in a similar manner.
228
JOURNAL OF GENERAL PSYCHOLOGY
the 25th session. Throughout the study body weights were recorded before each session, and S s had access to dry pellets and water for 1 hr following the experimental period.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
2.
Results
The mean body weight was one measure examined in the analysis of the results. From Figure 1 it seems that initially, for the seven days previous to adaptation, the mean body weight did not differ among the groups. The average body weight over these seven days did not differ reliably among the groups by analysis of variance (F = 2.13, df = 3/36, p > .05). From Figure 2 it appears that for the 15 days of adaptation, the mean body weight did not differ between the groups. By analysis of variance the average body weight from days 1-15 was not statistically different among the groups (F = 1.83, df = 3/36, p > .05). From Figure 3 it seems clear that from session 1-20 the 4-4 Ss weighed the same as the 4-16 group, the 16-16 Ss weighed the same as the 16-4 animals, and the 4-4 and 4-16 groups each weighed less than the 16-16 or 16-4 animals. The mean body weight from sessions 1-20 differed significantly among the four groups by analysis of variance (F = 5.16, df = 3/36, p < .01). By Duncan’s multiple comparisons the differences between 4-4 and 4-16, and between 16-16 and 16-4 were each not statistically significant (p > .05), while the difference between each shift group (4-16 and 16-4) and the respective control groups (16-16 and 4-4) was highly reliable (p < ,005). Four other measures were examined in the analysis of the results: average number of bar presses made during each session, average volume (ml) of liquid consumed during each session, average latency of the fifth bar press per session, average bar presses per min. Since average bar presses per min yielded essentially the same results as the above measures, excluding mean bar press latency, only average bar presses per min are reported here. Figure 4 is illustrative of three measures. From Figure 4 it is apparent that from sessions 1-19, 4% sucrose produced lower rates of bar pressing than 16% sucrose. The mean number of bar presses per min from sessions 1-19 differed significantly among the four groups by analysis of variance (F = 4.62, df = 3/36, p < .01). By Duncan’s multiple comparisons the differences between 4-4 and 4-16, and between 16-16 and 16-4 were each not statistically reliable (p > .05), while the difference between each shift group (4-16, and 16-4) and the respective controls was highly significant (p < .01).
LAWRENCE WEINSTEIN
.---.
.--+
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
O--U
4-4 16-4 16-16 4-16
DAYS MEAN BODYWEIGHTFOR IN
FIGURE 1 THE DAYSPREVIOUS TO ADAETATION, EXPERIMENT 1
229
230
JOURNAL OF GENERAL PSYCHOLOGY
24C
-E
0--0 4-4 M16-4 A-A 16-1 6 A-----A4-16
0,
W
I-
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
I
9 22c W 3
&
0 m Z
WQ 2oc I C
1
1
1
1
l
1
1
1
1
1
1
1
10
5
ADAPTATION DAYS
1
1
1
15
FIGURE 2 MEANBODYW E i C m FOR THE ADAPTATIONDAYS, IN EXPERIMENT I
Figure 4 indicates that on session 20 the shift groups abruptly changed their bar press rates to a level below their respective control groups (negative contrast effects). The mean number of bar presses per min during sessions 20-23 differed significantly among the two shift and two control groups by analysis of variance ( F = 4.62, df = 3/36,p < .01). By Duncan’s comparisons the differences between the shift groups and their respective control groups were each significant (p < .05). All measures except latency showed these significant concentration differences in the preshift and incentive contrast effects in the postshift. C.
EXPERIMENT I1
Experiment I indicated that a shift in incentive magnitude is confounded with a difference in body weight. In Experiment I1 saccharine solutions were
LAWRENCE WEINSTEIN
23 1
2404-4 16-4 =----I 16-16 o----o 4-16 +-a
230Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
lh
E rn W
220 -
=
"U
'0,
I
a-d
190
T 5 20
0 0
1 0
15
SESSIONS FIGURE 3 MEANBODYWEIGHTFOR THE EXPERIMENTAL SESSIONS WITH SUCROSE, I N EXPERIMENT I
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
232
JOURNAL O F GENERAL PSYCHOLOGY
0
5
10
15
20
25
SESSIONS used as incentives to examine the influence of a decrease in amount of reinforcement in the absence of :I confounded variable. 1.
Method
T h e S s were i 0 experimentally naive, male white rats. T h e details of age, housing, a n d apparatus were the same as in Experiment I.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
LAWRENCE WEINSTEIN
233
Levels of saccharine were selected from the studies previously cited (8, 11). O n the basis of a pilot study, it was determined that a 1 . 2 % saccharine solution produces a higher rate of bar pressing than a 1%, .lo%,and .01% saccharine solution. Ten Ss were randomly assigned to each of seven equal groups. Saccharine concentration treatmentsof 1 . 2 % , 1.0%,.lo%, and .Ol%,by weight, and tap water were used with control groups and four shift groups. The water (W-W), .lo% control group ( . l - . l ) , and .01% control group (.01-.01) were run throughout the experiment without changing concentrations. On the 15th session, the 1.2-.10 group was changed from 1 . 2 % saccharine to .lo%, the 1.2-.Ol Ss were shifted from 1 . 2 to .01%,the 1 - . 1 group was decreased from 1.0%to .lo%, and, finally, the .1-.01 group was shifted from . l o %to .01% saccharine. The details of adaptation, bar press training, and deprivation schedule were the same as in Experiment I except that performance appeared stable after 14 sessions. 2.
Results
The mean body weight was one measure examined in Experiment 11. Initially, for the seven days previous to adaptation, and for the 15 days of adaptation there was no statistically significant difference in body weight between the groups. From Figure 5 it seems clear that from sessions 1-15 all groups weighed the same. The mean body weight from sessions 1-15 did not differ significantly among the six groups by an analysis of variance (F = 2.10, df = 5/54,p> .05). The same four measures of Experiment I were examined in the present Experiment. Mean bar presses per min yielded essentially the same results as the other measures except average bar press latency. Only average bar presses per min are reported here. From Figure 6 it seems that from sessions 1-14 the 1 . 2 shift groups had the same rate of bar pressing, and the .10 shift and control groups reached the same level of behavior. Figure 6 further suggests that 1 . 2 % produced a higher rate of bar pressing than 1%, .lo%,and .Ol%;and 1.0%resulted in a higher level of behavior than . l o % saccharine which produced more bar pressing than . O l % saccharine. Furthermore, .01% appears to result in more bar presses per min than water. The mean number of bar presses per min from sessions 1-14 differed significantly among the seven groups by an analysis of variance (F = 2.34, df = 6/63, p < .05). By Duncan’s comparisons the differences between 1 . 2 - . I and 1.2-.01, and .l-.01 and .1-.1 were each not reliable (p > .05), while the differences between W-W and each other group,
JOURNALOF GENERAL PSYCHOLOGY
0--0 1*2- 0.1 .--. 1.2-001 &----A 1.0-0-1
A-----A04 - 001
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
0 - 4 04 - 0-1 [1--0 0.01-001
0
1
0
~
,
,
5
t
1
l
10
l
~
15
SESSIONS FIGURE 5 MEAN BODYWEIGHTI N GRAMSFOR THE EXPERIMENTAL SESSIONS WITH SACCHARINE, I N EXPERIMENT I1
1.2-.1 and 1-.1, andtheshiftgroups(1.2-.1,1.2-.01, 1-.I, and .l-.Ol)andthe respective controls were each reliable (p < .05). Figure 6 indicates that on session 15, groups 1.2-.1 and 1.2-.01 each suddenly decreased its bar press rates to a level below its respective control group (negative contrast effect), while the remaining shift groups reduced their bar press rates to the level of their respective control groups. The mean number of bar presses per min during sessions 15-20 differed significantly among the four shift and two control groups by analysis of variance ( F = 3 . 1 3 , df = 5/54, p < .05). By Duncan’s comparisons the difference between 1.2-.1 and . l - . I , and 1.2-.01 and .01-.01 were each significant (p < .05), while the differences between 1-.I and .1-.1, and .l-.01 and .01-.01 were each not statistically reliable (p > .OS).
l
1
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
LAWRENCE WEINSTEIN
5 MEANBARPRESSES
235
-
15
10 SESSIONS PER
FIGURE 6 MINUTEFOR SACCHARINE,
IN
EXPERIMENT 11
20
23 6
JOLRNAI. OF GENERAL PSYCHOI,O(;Y
All other mea3ures except average bar press latency indicated these significant concentration differences in the preshift and negative contrast effects in the postshift.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
D.
ESPERIMENT111
Experiment I1 indicated that negative contrast effects can be obtained with appropriate concentrations of saccharine solution which do not lead to weight differences. In Experiment 111 saccharine solutions were again used as reinforcement .05), while the difference between .01-1 and 1-1 was reliable Cp < .05).
E.
DISCUSSION
Experiment I clearly demonstrated that a change in incentive magnitude which results in negative and positive incentive contrast effects is confounded with a difference in body weight.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
LAWRENCE WEINSTEIN
23 9
The demonstration of negative contrast effects with a reduction in the concentration of a sucrose solution is not in accord with a number of studies in the literature (e.g., 5 , 7 , 10). The discrepant findings may be due to the different deprivation conditions selected by the present investigation and the three studies previously cited. In the present study, Ss were deprived of both food and water, while Dunham and Kilps, Guttman, and Rosen deprived their S s of either food or water. Quite possibly the present more severe deprivation schedule is more likely to produce negative contrast effects than food or water deprivation alone. In Experiment I the finding of positive contrast with an increase in incentive magnitude disagrees with many stuhes in the literature (e.g., 1,3,9). The discrepant findings, are, perhaps, accounted for by the more severe deprivation conditions employed in the present study as compared to the Ashida et al., Collier et al., and Metzger e t al. investigations. Experiment I1 clearly demonstrated negative contrast effects with a reduction in concentration of a constant volume saccharine solution. This result disagrees with Hulse (8) and Vogel et al. (11). As already discussed, the disparate findings are most likely due to the different preshift concentrations selected by the two studies previously cited and the present investigation. The finding that the water group responded a t a significantly lower level than all other groups indicates that the concentrations used were all reinforcers; none was a punishing agent (where a reinforcement is defined as an event which produces a significantly higher level of performance as compared to a control g r o u p i n this case the Ss exposed to water-which does not experience the event). Since negative contrast effects were not obtained with the saccharine concentrations employed by Hulse ( 1 .O and .lo’%) and Vogel et 01. (. 10 and .01%), the present contrast effects are quite probably due to the particular concentrations chosen, 1.2% and . O l % , and cannot be accounted for by differences in the present test situation and those situations used by Hulse or Vogel et al. Since negative contrast effects are obtained with saccharine solutions which do not lead to different body weights, the phenomenon cannot be explained on the basis of a confounding inherent in previous research with animals. The results of Experiment I11 indicate that positive incentive contrast effects are obtained with an increase in concentration of a saccharine solution. This result disagrees with Hulse (8).As already discussed, the divergent findings are very probably not due to different test situations, since, in the present study, contrast was not obtained with the concentrations used by Hulse; the
240
JOURNALOF GENERAL PSYCHOLOGY
contrasting findings are more probably due to the different preshift concentrations selected by Hulse and the current investigation. Since positive contrast effects are obtained with an increment in incentive magnitude in the relative absence of confounding factors, the phenomenon cannot be accounted for by adifference in body weight inherent in earlier animal research,
REFERENCES 1.
ASHIDA,S.. & BIRCH,D. The effects of incentive shift as a function of training. Psychor;.
2.
COLLIER,C. H . , & NOVELL,K . Saccharine as a sugar surrogate. 1.Comp. 6. Physio.!. Psyrhol., 1967, 64, 404-408. COLLIER,G . , KNARR,F. A , , & MARX. M. H . Some relations between the intensibe properties of the consummatory response and reinforcement. 1.Exprr. P s y r h o l . , 196 I . 62, 484-495. DILOLLO,V. D., & BEEZ,V . Negative contrast effect as function of magnitude of reward decrement. Psyrhon. S r i . , 1966, 5, 99-100. DUNHAM,P. J . , & KILPS,B . Shifk in magnitude of reinforcement: Confounded factors or contrast effects? J . Exper. Psychol., 1969, 79, 373-374. GONZALEZ,K. G., GLEITMEN, H . , & BITTERMAN, M. E. Some observations on the depression effect. J. Comp. Er Physiol. Psychol., 1962, 55, 578-581. GUTTMAN,N . Operant conditioning, extinction, and periodic reinforcement in relation to concentrationofsucroseusedasreinforcingagent.J. Exper. Psyrhol., 1953,46,2 13-222. HULSE,S. H. Partial reinforcement, continuous reinforcement, and reinforcement shift effects. J. Exprr. Psyrhol., 1962, 64, 451-459. METZCER,R . , COTTON,J. W . . & LEWIS,D. J . Effect of reinforcement magnitude and order of presentation of different magnitudes on runway behavior. J . Comp. Er Physiol.
Downloaded by [The University of Texas at Dallas] at 11:59 22 October 2014
S c i . , 1964, 1, 201-202.
3.
4.
5. 6.
7.
8. 9.
Psyrhol., 1957. 50, 184-188.
ROSEN.A. J. Incentive shift performance as a function of magnitude and number of sucrose rewards. J. Comp. Ci Physiol. Psyrhol., 1966, 62, 487-490. 11. VOGEL.J. K. MIKULKA, P. J. & SPEAR,N. E . Effects of shifh in sucrose and saccharine concentrations on ticking behavior in the rat. 1.Comp. Ci Physiol. Psyrhol.. 1968. 66, 10.
~
661-666.
WEINSTEIN. L. Magnitude of incentive contrast w a function of amount of verbal reward change. Psyrhon. S r i . , 1970, 21, 65-66, 13. . Negative incentive contrast with sucrose. Psychon. S c i . , 1970, 19, 13-14.
12.
~
clo John Bahchuk Box 1289 Vegreville, Alberta, Canada
