feedback on proposal

[gcohenfr]

You have a good analysis plan outlined in the proposal. However, your description is rather minimalistic and I can't understand from this proposal the challenges of the data and analysis. More specific comments follow:

• the table with independent variables is very clear. However, what is the size of the dataset? I've pointed this out in a previous issue to which you responded. These information should be included in the description of the data
• there was a full discussion on the distinction between public and private sectors. What was decided and why? from the proposal it seems that you will focus on the private sector. However, those decisions should be justified since you are going to analyze only a subset of the data. You need to justify this choice and even explain the limitations of your conclusions
• similarly, there was some discussion regarding variables related to managerial status. However, nothing is mentioned in the proposal about this variable. It seems that you are describing only the portion of the data that you will use (see previous comment for concerns on this approach)
• the research questions do not mention any interest in "prediction", however you aim for the best predictive model when selecting variables, why?
• it is interesting that you mention that you will "run model diagnostic". However, you don't state what you plan to do if the diagnostics fail.
• your variable selection plan is interesting and can help in selecting the most relevant variables.

[liuzhen529]

Thanks, Gabi! For the public and private sectors, we decided to do two ways as our client said before. One is to only focus on the employees in private sector; another is to use all data and figure out if public/private is a significant factor.

For managerial status, thank you so much for your suggestions. This variable is deleted because around 40% data are missing. We may think about it to see if we can do any further analysis on it.

Zhen