KwestResearch / engineering

Engineering Repositry
0 stars 0 forks source link

Interviewer Statistics Report Modifications #84

Closed mccuskk closed 8 years ago

mccuskk commented 8 years ago

Request for modifications to the Interviewer Statistics Report in Interviewer Management (not CATI)

Having used this Interviewer Statistics Report for several months could I request some modifications that may make it even better for me to use.

If you are happy with some or all of these suggestions, then you may be able to implement some of the easier modifications - such as items 1, 2 & 3 (the short term mod), plus possibly 4 & 5 - in under an hour of work. If so it would be great to have any such changes implemented within the next week. More substantial changes will just have to go on your work stack.

The main thrust of the suggested modifications is to make the reports simpler and therefore quicker to read. The suggested modifications are:

  1. Default display should be just for today (it is currently for 2 weeks)
  2. Don't display decimal points if a number is greater than 1.0. Display one decimal place if a number is is greater than zero and less than 1.
  3. Short term mod - Sort jobs in alphanumeric order by job number (g242, etc). Long term mod - ideally only display individual interviewer statistics for jobs selected in a drop down box as is done in Handheld (screenshot below shows individual jobs being selected in Handheld).
  4. Present the average figures for the overall list (i.e. all jobs) and for each individual job (currently all lists display each interviewer's figures but no average)
  5. Make the display of the second row for each interviewer (highlighted in yellow below) a selectable option (n.b. the second row displays time information).
  6. Create a new list displaying the average statistics for each job (i.e. not broken down by interviewer) - the Statistics equivalent of the Survey Summary shown in the screenshot below.
  7. Add two new columns after total numbers - percentage of time spent in interviews - percentage of time spend speaking with residents without securing an interview.
  8. Display the current columns from Called to Never Again in the following order (for clarity it might be worth leaving a space (e.g. empty column) between columns numbered 1, 2, and 3 below):

    Interview Never again Call Again Not this time Not Applicable Interview Abandoned Resume later

    Wrong number Dead line No 141

    Called No reply Answer phone

    Lang Unavail (better to remove this from table) Call another (better to remove this from table) Misdial (better to remove this from table)

After these proposed modifications the Statistics tables will still contain a forest of numbers. So, if the effort involved was not too large it would be really valuable to flag up any percentages that are outside set limits - for example if 'Never again' exceeded 1%. The most obvious flag might be to colour the cell yellow. You might want to test the feasibility of 'flagging' and if you think it isn't too difficult then I could suggest some possible limits for a few of these statistics.

mccuskk commented 8 years ago

Initial commit of this work.

I have implemented items 1,2,5 and 8. Item 3 is how the system currently works.

I have modified item 8 to remove called from list 3 as it is not really relevant when using dialer and moved lang unavailable to the end of list two.