الفهرس 
الفصل الاوليوجد فقط 14 صفحة متاحة للعرض العام
 |  | from | 274 |  |  |
| | | from | 274 | | |
المستخلص
INTRODUCTION:
The basis for multiple cue probability learning (MCPL) research is the belief that an individual’s decision or prediction is a function of probabilistic events which occur within his environment.Even though cues may provide only partial information about the environment (ecology), individuals still must rely upon them to varying degrees in arriving at a decision.
Brunswik (1952) was one of the earliest proponents of the study of probabilistic situations and of determining the manner in which individuals handle these probabilistic relationships.He formalized the lens model as a means to conceptualize and analyze probabilistic relationships between the individual and his environment. STATEMENT OF THE PROBLEM:
Review of the literature in the field of probability learning showed that previous studies did not deal with the variables of the individual learner,especially reasoning ability and tolerance of ambiguity.It is clear that these variables are very important in any learning situation.
The present study proposes that individuals may do inference processing when performing a learning task,also that the degree of the individual’s tolerance of ambiguity interacts with any situation characterized by ambiguity in its probability nature.
The present study,therefore,deals with the effect of reasoning ability and the tolerande of ambiguity.The effect of a third variable,cue validity,is also examined,with particular reference to the consequences of holding redundancy and predictability constant. HYPOTHESES:
The present study tests the following hypotheses:
1-Achievement,consistency and matching as measures of learning prediction behaviour differ according to the subject’s level of reasoning ability.
2-Achievement,consistency and matching as measures of learning prediction behaviour differ according to the subject’s level of ambiguity tolerance.
3-Achievement,consistency and matching as measures of learning prediction behaviour differ according to the level of cue-validity.
4-Achievement,consistency and matching as measures of learning prediction behaviour differ according to the interaction between reasoning ability and ambiguity tolerance.
5-Achievement,consistency and matching as measures of learning prediction behaviour differ according to the interaction between reasoning ability and cue validity.
6-Achievement,consistency and matching as measures of learning prediction behaviour differ according to the interaction between ambiguity tolerance and cue validity.
7- Achievement,consistency and matching as measures of learning prediction behaviour differ according to the interaction between reasoning ability and ambiguity tolerance and cue validity respectively. PROCEDURES: 1-TOOLS: A)REASONING ABILITY TEST
Reasoning ability is defined in terms of conditional reasoning because the present study suggests that probabilistic tasks require high conditional reasoning ability if correct criterion inferences are to be made.Conditional reasoning is defined as ”A hypothetical construct which can appear in performance in conditional sentences”.
The conditional reasoning test consists of 36 items with a reliability of .73 (p<.01) by Spearman Brown and a validity of .51 (p<.01) when correlated with another
general reasoning test. B)TOLERANCE OF AMBIGUITY SCALE
Tolerance of ambiguity is a tendency to perceive or interpret as acceptable or desirable information that is marked by vague,incomplete,fragmented,multiple,probable, contradictory,or unclear meanings,and intolerance of ambiguity is a tendency to perceive such information as actual or potential sources of psychological discomfort or
threat.
The tolerance of ambiguity scale consists of 40 items with a reliability of.68 (p<.01) by the split half method,and of .54 (P<.01) by Alpha Cronbach.Internal consistency is also high.There are 8 factors resulting from factor
analysis of the scale. C) GENERATING OF CUES AND CRITERION VALUES:
Cues and criterion values were generated by using a special program for the computer to obtain the desired relationships among the cues and criterion in each block of trials.The correlation coefficients between cues and criterion which represent the cue validity were (+.5 and +.8) and (+.5 and -.8) for both treatments.The redundancy (the correlation between the cues themselves) held constant at .00 in both treatments,also the predictability was .90 and constant. 2-EXPERIMENTAL DESIGN:
The experimental design was factorial design 2 (two levels of reasoning ability) X 2 (two levels of ambiguity tolerance) X 2 (two levels of cue validity) X 10 trial blocks with repeated measurement on the last factor. 3-SAMPLE:
The sample for the study consisted of 112 students from Benha College of Education.This sample was selected after administering the Conditional Reasoning Ability Test and Ambiguity Tolerance Scale.Ss with high scores and low scores in the Conditional Reasoning Test and in Ambiguity
Tolerance Scale were selected.The Ss were assigned to 8 experimental groups,with 14 in each group. 4-EXPERIMENTAL PROCEDURES:
The values of cues and criterion were typed on a continuous roll of paper which was fed through an over-head projector.The projector was used for displaying cues and criterion values as columns representing the values.On each learning trial,the subjects in each experimental group were instructed to:
1-Observe the two cue values represented by the columns 2-Record their prediction response on their answer sheet 3-Compare their prediction response with the criterion value displayed on the projector for outcome feedback. There were 250 trials for each experimental group divided
into blocks of 25 trials. 5-DATA ANALYSIS:
There were three performance indices,namely achievement ,consistency and matching computed for every subject in each block of trials.These indices were transformed by Fisher’s transformation to Z values,and then analyzed by
2 X 2 X 2 X 10 analysis of variance with repeated measurement on the last factor. RESULTS OF THE STUDY:
The results of ANOVA 2 X 2 X 2 X 10 with repeated measurement for the three indices Achievement,Consistency and Matching were as follows:
(1)High Resining ability Ss were significantly higher than low reasoning ability Ss in achievement and consistency but not in matching.Therefore the first hypothesis of the study was supported in achievement and consistency only.
(2)There were no significant differences between the tolerant of ambiguity Ss and intolerant Ss in the three indices,achievement,consistency and matching.Therefore the second hypothesis was not supported in any of the
performance measure.
(3)The third hypothesis was supported in the three performance measures as there were significant differences between the valid cues condition and invalid cues condition in favour of valid cues in achievement,consistency and matching.
(4)All the hypotheses of the interactions among the main variables: conditional reasoning ability ,tolerance of ambiguity and cue validity were not significant in thethree performance measures,achievement,consistency and matching. Therefore,for example,there. was no significance for the interaction between conditional reasoning ability and cue validity. DISCUSSION OF RESULTS
The results of the study are discussed in the light of the theoretical framework of the study and in terms of previous research findings.Further research is suggested for exploring the effects of the individual’s traits and abilities in these probabilistic inference tasks.