Hi everyone. Could you please answer to me ASAP? I have questions on chapter 12. 1. In the grouping of signs test, could you tell me how i can determine the number of groups of positive deviations? (example 2.Almost all exercises in the chi-squared test have to do with graduation by reference to a standard table. If we are not given an equation and we are only told that there was a reference with a standard table, how many degrees of freedom should we deduct? or should we not deduct at all? I am confused as in some past papers we deduct for example 2 or 3 and in some others 0. 3. In April 2017 question 10, i cannot understand how we conducted a chi squared test as we are only given the ranges and not the exact values of Zx. Could i take another test rather than chi squared? 4. In September 2015 Question 11 i cannot understand the rationale behind the answer of IV. 5. Also, in the chi squared test we have (actual - expected)^2/expected. Could you please explain to me why the graduated rates are the expected ones? Thank you everyone!
1) Take September 2011,Q10 for example, Here, we have 2 groups(or runs) of positive deviations, ie. One comprising of the term 0.988 and the other group going from 0.026 to 0.880 2) If the graduation is performed with reference to a standard table and no parameters are fitted, then we reduce some degree of freedom, usually 2(sometimes 1 or 3 also depending on the sample size). In questions where no degree of freedom is lost, we were not given graduated rates, rather just asked to comment on the observed and expected rates. 3) The examiner's comments in the solution states where can this test be found in the material. Also we are asked for 2 tests, so we can perform a test for outliers and another test for bias. 4) We are asked to find the impact on variance by the given duplicate policies scenario. We have just applied the formula given which shows how the variance increases in such cases. 5) Because we expect our crude rates to be equal to our graduates rates
