Correlations Between Factors of Confidence, Accuracy and Time - Essay Example

?Running head: CORRELATIONS BETWEEN FACTORS Correlations Between Factors of Confidence, Accuracy and Time College Affiliation Results of this report indicate that confidence influences accuracy and that time affects accuracy depending on the level of confidence. The majority of participants in the analysis were of female gender. One sample t test indicated that the average accuracy level is not equal to the confidence level. The ANOVA test indicated that there is a relationship between accuracy and confidence and the p value is 0.000. The relationship is identified using the average values across the confidence level and as confidence increases, then accuracy also increases. The time factor does not improve the relationship between confidence and accuracy. It is possible that time influence's a person's judgment and correlates with confidence and accuracy. However, time affects accuracy at different confidence level. At a 55 percent confidence level, time is likely to improve accuracy; however, at 95 percent confidence level, time does not improve accuracy. Correlations Between Factors of Confidence, Accuracy and Time Previous studies indicate that there is a relationship between confidence and accuracy. Scholars argue higher confidence levels result in higher accuracy levels. However, this relationship may not hold true if individuals exude an over-confident attitude. This paper focuses on highlighting the relationship between confidence and accuracy and also explores the relationship between time and accuracy. Some scholars argue that there is a relationship between time and accuracy due to individuals being given time to mentally process a question before giving an answer; therefore, they are more likely able to provide a more accurate answer. The main focus in this paper is to identify the relationship between accuracy and confidence and how it correlates with time as a factor. The relationship between time and accuracy is also evaluated in and if it actually is a factor in an improvement of the relationship between accuracy and confidence. This report is drawn from previously obtained research in order to test the possibility that time and accuracy have an impact on judgment in conjunction with confidence levels. Secondary data collected in the past are also used in this analysis. Possible limitations in this analysis include the fact that data may not contain other variables that may impact the outcome of the conclusions of these hypotheses and may additionally be useful in exploring how these three factors relate to one another, if at all. Literature Review According to Allwood and Jonsson. (2003), defines confidence as how sure an individual is of themselves and in this experiment, is in regards to answering a question given and how the answer will depend on his or her own knowledge and beliefs. In his research, he assumes that individuals are likely to rate their confidence with reference to their accuracy. If they are confident, then they are sure they are not wrong. Therefore. if individuals rate that they are 50 percent confidence, then they also anticipate to be 50 percent accurate. Krug, (2007) on the discussion of the use of eye-witness testimony, states that confidence can be a good indicator of accuracy. Additionally, other authors such as Safer and Wise (2004) state that many courts in the United States will depend on a witness' confidence to determine whether or not the information they provide is accurate. This mentality is supported by Bradfield and Wells, (2000) who determined that the jurors' opinions highly depended on their confidence to make decisions. According to their research study, over 70 percent of prosecutors and law enforcement officers depended on witness confidence to measure the accuracy of information provided, therefore determining if the answers they are giving are in fact true. In contrast, many authors such as Memon, (2001) state that confidence is a poor measure of accuracy and that there is only measured to be a 0.07 correlation between confidence and accuracy in their study. Rishworth, (2002) however, states that confidence actually can be a good measure of accuracy, providing that all the factors are taken into consideration. Some of the factors the author highlighted include the contrast of the characteristics of overconfidence and under-confidence. Olsson, (2000) highlighted the hard-easy effect, and according to description, this effect refers to a situation where when accuracy is high, then overconfidence tends to be low,. In retrospect, when accuracy is low, overconfidence tends to be high. However, Allwood et al (2003) argues that domain of overconfidence is not yet agreed upon. For example, when overconfidence is likely to be experienced by an individual, this means that this effect is not easy to identify. Allwood et al (2003) also highlighted this same effect, focusing on the factor of overconfidence. According to this author's research, when individuals are overconfident, this then means that their confidence level is actually greater than their accuracy level. This paper aims at testing the relationship between accuracy and confidence and also how time for deliberation plays a role in the relationship between the two other measures. In order to analyze this relationship, the following hypotheses are tested: hypothesis one: confidence is equal to accuracy; hypothesis two: confidence leads to accuracy; hypothesis three: extra time improves accuracy; and hypothesis four: time improves the relationship between accuracy and confidence. Method Participants A total of 86 participants were included in this experimental analysis. Seventy-two percent of the participants were female while 28 percent were male participants in this sample group. The average age of the participants was 22 years of age with a standard deviation of 5.119 years. Any other variables such as race, background, religion, ethnicity, motivation levels, their environmental conditions, etc. were not taken into account in regards to what other factors may contribute to their responses. This sample was a subset of previous data from 2010 and manipulated to control for confounding variables. Design The sample for this experiment included a series of one large focus group that was then subdivided into two equal groups. Each of the test groups contained 43 participants each and were labeled as Group One and Group Two. These two groups were given questionnaires displayed on a Powerpoint slide show to analyze and provide answers for. Group One was slower and took more time to deliberate the questionnaire assessment, allowing more time for them to go through the judgmental process for a possible solution to the questions. Participants of group two were faster and took less time to analyze the assessment. Following the questionnaire assessment, participants were then asked to rate their confidence upon completion of answering the questions of the assessment. Assessments were rated with reference to the confidence of the participants and they include a 55 percent confidence where participants stated that they guessed the answer. Sixty-five percent of participants stated that it was almost guess, 75 percent were unsure, 85 percent were almost certain and 95 percent were for certain. Results After determining the confidence level results, 48.74 percent of the participants that guessed were actually correct in their answers of the questionnaire. Of the participants that said they had an almost guess, 52.67 percent were correct. The participants that claimed they were unsure, 54.80 percent were correct. Those that were almost certain, 62.45 percent were correct and the participants that were over-confident and were certain they were correct, 69.20 percent of them actually were. Additionally, the following are the results that were determined using SPSS analysis. Gender Participants' gender was analyzed and the distribution determined. Results indicate that the majority of the participants were female, representing 72 percent of the group and 28 percent of the participants were male. It is undetermined whether or not their gender had any affect on the actual outcome of their accuracy. Age The age of the participants was also analyzed and the following were the descriptive statistics: Statistics Age n Valid 86 Missing 0 Mean 22.09 Median 20.00 Mode 19 Std. Deviation 5.119 Results indicate that the average age of the participants was 22.09 years, the median was 20 years and the mode was 19 years. Standard deviation from the average value was 5.119 years. Results of the relationship between time, accuracy and confidence may have been different if the test would have been given in a different environment with a more variance in age. Groups and accuracy mean The average accuracy values for each group with reference to the confidence level were analyzed. Confidence levels are represented by a: guess; b: almost guess; c: unsure; d: almost certain; and e: certain. The following table displays the results: Group (Slow or Fast) Measure of Confidence Measure:MEASURE_1 Group (Slow or Fast) Confidence Mean Group One (Slow) a 47.405 b 53.342 c 54.602 d 62.172 e 70.170 Group Two (Fast) a 50.081 b 51.988 c 54.998 d 62.730 e 68.228 Group one had higher average values of 55 percent, 65 percent and 95 percent confidence levels, while group two's average was higher for the other confidence levels, indicating more certainty that they were accurate. Hypotheses Four hypotheses were tested and they are as follows: Hypothesis One: Confidence is equal to accuracy. Hypothesis Two: Confidence leads to accuracy. Hypothesis Three: Extra time improves accuracy. Hypothesis Four: Time improves the relationship between accuracy and confidence. To test these hypotheses, ANOVA and t tests were used and the results are as follows: Hypothesis One: Confidence is equal to accuracy Confidence levels were identified as G =55 percent, AG =65 percent, U=75 percent, AC=85 percent and C= 95 percent, with the main aim to test whether or not the accuracy level is equal to these confidence levels. The following are the average accuracy values for each confidence levels: Average Values for Confidence Levels Measure:MEASURE_1 confidence Mean Std. Error 95% Confidence Interval Lower Bound Upper Bound 1 48.743 1.180 46.396 51.090 2 52.665 1.695 49.294 56.037 3 54.800 1.605 51.608 57.992 4 62.451 1.528 59.413 65.489 5 69.199 1.417 66.381 72.016 The above results indicate that accuracy levels for G = 48.743, AG = 52.665, U = 54.800, AC = 62.451 and C = 69.199. The following t tests were given: H0: G = 55 percent and H1: G Read More