The Effect of Two Competitive Conditions and a Non-Competitive Condition on Performance and Retention on a Recognition Task

Based on a review of the literature on competition, several hypotheses on the effects of competitive instructional set on both speed and accuracy measures were explored usin g a paired associate learning paradi gm, One hundred and forty second graders (70 boys and 70 girls) were randomly assit;ned to three instructional treatments: speed riompetition, accuracy competition and non-competition. The competitive groups were further divided into relative positions of win, lose and tie; subjects were equally divided by sex, Althou gh the subjects in the speed competition group performed significantly faster than subjects in the other conditions, no significant differences were found between any of the treatment conditions and the non-competitive condition on number of errors in performance, Indications were that males and females may employ a different strategy approach to noncompetitive situations but that both respond equally to a competitive situation. Several possible explanations for these results are discussed, The implications of these results for education indicate that care and caution should be used in applying co~petitive instructional sets as a un{versal classroom technique.


THE EFFECT OF TWO COMPETITIVE CONDITIONS
(1) Competitive instructional set is more effective in a ss i s ti ng the learning of recognition tasks than is a non-compet i ti ve instructional set, (2) Competitive instructional set is more effective in assistin g the learnin g of recognition tasks which emphasize speed rather than accuracy.
(3) Competitive instructional set enhances retention more than does non-competition, .
(4) Competitive instructional set enhances retention when speed is emphasized more than accuracy, (1) a, The competitive groups will make fewer errors than the non-competitive group.
b, The competitive groups will perform faster than the non-competitive group.
(2) a, The speed competition group will make fewer errors than the accuracy competition group.
b, The speed competition group will perform faster than the accuracy competition group, (J) The competitive groups will make fewer errors on the retention tasks than does the non-competitive group, (4) The speed competition group will make fewer errors than the accuracy competition group on the retention tasks, (5) Subjects in the tied condition will make fewer errors than those in the other two positions, four because they did not understand .the task, thirteen because the task proved too easy for them. Those eliminated were evenly divided between males and females (one more female was dropped than males) and were spread across all treatment conditions~ No discernible pattern as far as IQ, SES or achievement was found.

Assignment to Groups
One hundred and forty subjects were assigned to treatment conditions, according to the following design,

Figure 1. Schematic representation of treatment conditions
Subjects were assigned to one of three conditions, speed competition, accuracy competition .and non-competition.
The speed competition condition emphasized answering correctly first, the accuracy competition condition emphasized the total number right, and the non-competitive group was simply asked 12 to perform the task with no reference to a comparative group.

Procedure
The equipment was placed in a 20 x 10-foot room of a mobile office which was then moved from school to school, Subjects were seen individually by one of the experimenters, either a male or a female, in a counterbalanced manner controlling for sex of experimenter.
In order to acquaint the subject with the task, all participants went through a trainin g phase in which they were asked to learn a practice list of four pairs of pictures mounted on J x 5-inch index cards and slides, Each pair was presented followed by a card containing a single .stimulus picture.
The following instructions were reads This is a task to see how well you can remember pictures. First, you'll . see a pair of pictures like this (illustration pair); then, you'll see only the picture on this side (illustration stimulus card), Your job is to guess what goes here, To help you guess, here are the pictur e s of the answers (point to pictures on station), All you have to do is to push the button in front of the picture that you think goes here, W hen you do, one of these lights will li ght up to sho w you which one you picked, W hich one would you push for this (illustration)?
OK, and this lit up to .show you which one you picked, Now these li ghts here will show you whether you are right (turn on green, yes) or wron~ (turn on red, no), In this case, you're right (yes). Let's try these, OK?
The subject was then presented with the four practice pair s and the stimulus cards to which he/ s he was asked to re s pon d .
The experimenter continued with the following instructions: Now that you have the idea, I would like to continue to practice with these pictures -onl y you'll now see the pictures up here on the screen, First, you will see the four pairs; then you will see only the picture on this side. Push the button in front of the picture you think goes ·with the one up on the screen.
At this point, any subject who did not get at least one ri ght was replaced by a student from the alternative list. After the practice trial was run, the subject was further instruct ed: OK, now that youi~e practiced, I will show you a new set of pictures on the screen. First, you will see the pairs and then just the stimulus picture, Ail you do is push the button in front of the picture you think goes with it.
Then, depending on treatment condition, the following instr uctions were given: Control: W hen you are finished with each set or trial, a card like this will be put in front of you to show you that a new set is coming up, So, a card like this would mean that this is the second trial,

Speed Conditions
This has already been done .at another school, and I would like to see how well you do compared to another child at that other school, The machine already has that other child's answers and how quickly that child did them, I would like to see how quickly you can get the right answers compared to that child, To help you know how well you're doin g , after each set of pictures, I will place a card like this in front of you.
It will show you whether you are winning, losing or tied, So, if you see a card like this, it means that out of all the ones you've done so far, the other person has gotten four faster than you have, This would mean that you have gotten four faster, Now, what would this mean? Remember, the score tells you whether you are getting the answers faster than the other child, You will know if they are right or wrong by these lights. Any questions?

Accuracy Conditions
This has already been don e at another school, and I would like to see how well you do compared to another child at that other school, The machine already has that other child's answers, I would like to see how m~ny you can get right compared to that other child, To help you know how well you are doin g , after each set of pictures, I will plate a card like this in front of you, It will show you whether you are winnin g , losin g or tied, So, if you see a card like this, it would mean that out of all the ones you've done so far, the other person has four more ri ght than you do, This would mean that you would have four more right than . 4) *p     Table 2.

Levels
Mean latencies to respond for three speed competition groups (sp), three accuracy competition groups (ace), and a non-competitive group (nc) A 2 (condition) x 3 (level) x 2 (sex) factorial analysis (W iner, 1971) was applied to the data on latencies, and the results are reported in Table 4. There is a significant main effect found between the two groups on condition with the speed competition (sp) group s taking significantly less time to respond than the accuracy (ace) groups (X (sp)=120 , 9; X (acc)=14?,1).
Since there was a correlation between the "no response" category and latencies, an analysis of covariance (W iner, 1971) was performed, using "no response" as the covariate, These results are reported in Table 5, Wh en "no response" is used as a covariate, the difference between the groups in terms of the speed at which they perform the task remains statistically significant, Therefore, the time difference in performance cannot be simply attributed to the use of the "no response" category, . A comparison was then made between each of the experimental groups and the non-competitive group, using Dunnett•s t statisti6 (Winer, 1971), Results are reported in Table 6, Source   In summary, th 'e speed competition group is faster than the accuracy competition and the non-competitive groups, but there were no significant differences between any of the groups in the number of errors associated with performance nor on the immediate test, There was a significant difference between the non-competitive group and the competitive groiJps on the use of "no response" as a strategy, and the speed competition group performed the task significantly faster than the other groups without any significant difference in errors, ---------------------  14.65 6,0J