when is a sign test appropriate?

James Smithson

New member
Joined
Nov 6, 2020
Messages
45
I have a table, these figures are estimates, so there is some degree of uncertainty (or error) associated with the values. For example, there could be sampling error or response bias in the survey data used to produce these estimates.

Place Estimated income (£)
England 20 988

North East 17 809
North West 18 861
Yorkshire 18 119
East Midlands 18 932
West Midlands 18 885
East of England 22 081
London 29 825
South East 24 568
South West 20 984


someone reads in a different report that the population median income for Englan was £18 500. Considering the data in Table, does this value seem plausible to you? Justify your answer.


So far I have
worked out the test statistic is 2 (as its the lower number when working out figures above and below 18 500.

and used a sign test to evaluate the hypothesis that the population median income in the nine English regions is 18 500

I worked out the appropriate critical value at the 5% and decided I can not reject the hypothesis. I have also worked out the pvalue that also says i cant reject it

However where I am really struggling is a conclusion I am not sure how to conclue the answer , I know my sign test and p test state we can't reject the hypothesis but I feel the hypothesis has to be wrong because the median is clearly higher than 16500 by some distance . In this conclusion I really want to talk about weather ths sign test is appropiate or not and i want to make specific references to the context .


( I worked out p value of 0.180 and the critical value at the 5% significance level to be 1 when there are 9 data values )


Any advice is appricated as its all new to me

thanks in advance James
 
Top