Variables in the Test Set for Item 266510

[mathgie]

Hello!

A quick clarification about item 266510:

In the "Example_test_data" file we gave, it appeared as though the variables named "eliminated," "eliminations," and "selected" all have NA for these variables. In actuality, these will be included in the test dataset; however, they are very sparsely populated. For “eliminated” and “eliminations”, about 95% of these are NA and only about 5% have meaningful data. For the “selected” variable, only about 40% of this variable contained a vector of responses in the “FALSE TRUE FALSE FALSE” – style format with the rest being either "NA" or “FALSE FALSE FALSE FALSE”.

[czopluoglu]

I appreciate this clarification. I have a related question.

The statement in the readme.pdf leads to an assumption that any NAs in the example test data indicate suppression of those variables for a given item in the test set that will be released later. For instance, if you look at the example test data, variable accom2 is also set to NA for Item VH304954. Is it meant to be set as NA because it will be suppressed in the test set? Or is it also another sampling error when you create the example_test_set?

Similarly, pta_rtr1 is set to NA for some items and not for others. Can you please confirm which items pta_rtr1 will be suppressed in the test set?

Thank you.

[mathgie]

Hi Cengiz,

In general, you will have all of the demographic information that you see in the training set also provided for you in the test set. If you look at the codebook, it says: "Note: Item VH304954 did not have accom2 so for this item accom2 is entirely NA." This is why this demographic variable is empty just for this item. It is empty for both the training and test set.

In looking at the other variables, you will get pta_rtr1 for: 266015, 269384, 271613, and 507804. You will get ptb_rtr1 and ptc_rtr1 for 302907.

Please let me know if you need any additional information or if this wasn't clear. The rest of the test set follows the examples in the example_test_data sheet.

[czopluoglu]

Great! Thank you for all the information and clarification.