Professional Documents
Culture Documents
Stats Assignment 1
[Type the document subtitle]
Robert Petzer Nathanael Seeber Tracy Favish Andrew Reeves Clayton Redford Bradley Tighy Bradley Campleman
448889 (Finance) 518599 (Finance) 481059 (Finance) 304092 (Finance) 389261 (Finance) 376372 (Finance) 435792 (Finance)
Question 1 The data used is randomly selected. We made use of both gold and platinum share prices from various companies. The values are completely random and no real world sample was used. Therefore there is variability in the data. If, for example, we used the actual historical data on these companies, there would be less variability. This variability is also affected by the method used to collect the data as well as the sample used. A completely random sample results in large variability. In the real world, gold and platimun share prices are affected by investor sentiment, political upheaval, public information and various macroeconomic factors such as the interest rate, inflation etc.
Question 2
149.482500029.899433398.6600000206.9800000
Question 3 a)
b)
c) The mean, median, mode and inter-quartile range will remain the same as the data does not change. However drastic changes occur in the histograms distribution. The second histogram showed a slight skewing of the data to the left. The increased number of intervals makes the graph more detailed and easier to interpret. No changes occured in the box-and-whisker plot. The box itself represents the inter-quartile range.
Question 4 A hypothesis could be formulated to test the mean share price before the Marikana tragedy. Based on the above data, we will test whether the mean share price is significantly different from R120. We will test at a significance level of 5%. H0: =120 H1: =120
Question 5 a) H0: =120 H1: =120 The mean price of shares before the Marikana tragedy is R120. (Null Hypothesis) The mean price of shares before the Marikana tragedy is not R120. (Alternative Hypothesis) Two-tail test.
t Test
The TTEST Procedure Variable: BeforeMarikana N Mean Std Dev Std Err Minimum Maximum
207.0
11
3.420.0058
The T-stat exceeds the T-critical value (3.42 > 2.201) and the P-value is less than the significance level (0.0058 <0.05) Therefore we reject the null hypothesis at a 5% level of significance and as such the mean share price before the Marikana tragedy is significantly different from R120.
b)
t Test
The TTEST Procedure Variable: BeforeMarikana N Mean Std Dev Std Err Minimum Maximum
207.0
11
3.420.0058
No, it does not make a difference. Since the P-value of 0.0058 is still less than 0.01, we reject H0 at 1% level of significance.
c) The P-value for the 2 sided test = 0.0058 Therefore, the P-value for the one-sided test is 0.0029. This is less than both significance levels of 0.05 and 0.01. d) Our conclusion would change from reject to do not reject for the two-sided test at a level of significance of 0.58%.
Question 6
t Test
The TTEST Procedure Variable: BeforeMarikana Commodity N Mean Std Dev Std Err Minimum Maximum
207.0 187.4
G 148.4 121.3 175.5 32.9275 22.1289 68.7940 P 150.6 126.2 175.0 29.6584 19.9319 61.9640 Diff (1-2) Pooled -2.2017 -34.9920 30.5887 31.3356 23.1595 49.9199 Diff (1-2) Satterthwaite -2.2017 -35.0282 30.6248
Method Variances DF t Value Pr > |t|
Folded F
a) H0: 1 = 2 H1: 1 = 2
1.230.8241
The null hypothesis states that the mean share price of gold is equal to the mean share price of platinum. The alternative hypothesis states that the mean share price of gold is not equal to the mean share price of platinum. The F-test for the equality of variances gave us a p-value of 0.8241 which is greater than 0.1. Thus, at a 10% level of significance we fail to reject H0: that the variances are the same.
Since we failed to reject the null hypothesis, we used the pooled test to test the hypotheses again. This gives us a P-value of 0.9056. Since this is still greater than the 10% level of significance we failed to reject the null hypothesis. This shows that the mean of group 1 is equal to the mean of group 2. b) The 90% confidence interval using the pooled method is (-34.9920;30.5887) c) Testing at a 1% level of significance makes no difference to our findings as the p-value of 0.9056 is greater than 0.01. d) It would change at level of significance equal to 90.56%
e)
t Test
The TTEST Procedure Variable: BeforeMarikana Commodity N Mean Std Dev Std Err Minimum Maximum
207.0 187.4
G 148.4 94.1793 202.6 32.9275 17.9904 114.7 P 150.6 101.8 199.4 29.6584 16.2043 103.4 Diff (1-2) Pooled -2.2017 -59.5389 55.1356 31.3356 19.7442 67.4882 Diff (1-2) Satterthwaite -2.2017 -59.6786 55.2753
Method Variances DF t Value Pr > |t|
Folded F
1.230.8241
Question 7 a)
t Test
The TTEST Procedure Difference: BeforeMarikana - AfterMarikana N Mean Std Dev Std Err Minimum Maximum
19.826710.968328.685017.086912.776226.4956
DF t Value Pr > |t|
11
H0: D = 0 i.e 1 - 2 = 0 H1: D = 0
4.020.0020
The p-value of 0.002 is less than the 3 significance levels of 10%, 5% and 1%. Thus at all levels of significance we reject the null hypothesis and find that the means for each sample are statistically different i.e. the mean share price changes post-Marikana. b) There is no difference in our conclusion at the different significance levels as 0.002 is less than 0.1, 0.05 and 0.01. We reject the null hypothesis at all levels of significance. c) Our conclusion would change at a significance level of 0.2% - at this level we would fail to reject the null hypothesis.