Question

Three hundred people working in 40 state offices process X120 forms To evaluate a proposed change in the X120 form, would you randomly assign offices or individuals? That is, if offices are assigned, 15 offices would use the form with the proposed change, and 15 offices would continue to use the existing form. If individu- als are assigned, roughly 50 percent of the individuals in the office would use the form a with the proposed change, and 50 percent would continue with the existing form. b Would you allow subjects to volunteer to test the new X120 form? Justify your answer c Would you test the new X120 form in the three slowest offices and evaluate the change in productivity? Justify your answer

Answer 1

Note: This response is in UK English, please paste the response to MS Word and you should be able to spot discrepancies easily. You may elaborate the answer based on personal views or your classwork if necessary.

(Answer) (1) In order to evaluate a new variable (the new x120 form), there would have to be a control group and an experimental group. This means that a part of the sample would have to adopt the new method and a part of the sample would have to retain the old method. The difference and success rate between the two can be then compared in order to measure the efficiency.

If, on the other hand, all offices were to adopt the new method, it would be essential to compare the new results with the results of the old method by the sample size of one week or month. In this way, the results of both months with different x120 forms can be compared to see which one is successful.

(2) There are several pre-testing and post-testing methods that help pinpoint problems before an upgrade is fully adopted. If there were volunteers who tested the new method and were interviewed about their thought, it would be possible to make any changes or streamline the process wherever necessary. Therefore, it is always better to adopt the “prevention is better than cure” ideology when adopting such new techniques.

(3) Yes, the test in the three slowest offices would be of a great value. However, the results must be compared with an office of a normal speed. This would avoid the problem of statistical regression in internal validity that arises from taking only the extreme variables into consideration. With offices of mixed speed, it would be easier to accurately measure whether the x120 functions well under different conditions.