Answer

1) a). The engineers may be taken as a homogeneous group and since the ministry believes that the comments depend on the engineers’ category, a stratified sampling method would be recommended. An equal composition of each category may be first selected to have a sample of 100 and then send a questionnaire through e-mail. This is due to the fact that the results are needed urgently.
Assumptions
* Population of engineers are homogeneous population of each type of engineers separately * Engineers are normally checking e-mails, and the questionnaires will be responded in time.

b). The two distributions are found to be positively and negatively skewed respectively Age: - slightly positively skewness 0.31 OT :- Negatively skewned -1.36 Therefore the best measre of central tendency is median and best measure of dispersion is Inter Quartile Range or Quartile Deviation

2014 Question No 2

Answer

2 a)
i) Let µu = Mean rate of number of production by a worker per hour under usual methods. µv = Mean rate of number of production by a worker per hour under Viskam’s method. Then, H0- The rate of production of a worker per hour, has not increased with Viskam’s methods H1- The rate of production has increased with Viskam’s methods.

ii) A pared T-Test may be recommended. iii) The samples are normal, related and are with equal variance. iv) If the variances are not equal conduct the same test without taking variance as equal. If the two samples are not normal, then a non-parametric willcoxon test may be carried out. v) The p-value is greater than the level of significance. 0.82 > 0.05
Hence we cannot reject Ho. The rate of production of workers per hour has not improved by introducing the Viskam’s method.

b) Standard error is;
When you take a point estimate through a sample, it is accepted that this estimate