Explaining event and student logging

[perspegrity5]

Yvonne from WestEd is asking about some of the data columns in the StudentProblemHistory table. I have answered many of her questions but need your help with some. IVON: you may know these answers also--please comment!

1. What units are these variables in? timeInTutor timeInSession timeToFirstAttempt, timeToFirstHint, and timeToSolve

Can you check the code and answer this question? It might be milliseconds for some, and in seconds for others.

2. Is the Problem History Log calculated/derived/aggregated up from the Event Log? Tom's answer: No it is not. Entries in the table are added directly in MS code; i.e. entries in the PH table are not derived from processing the EventLog data table. Though, if the code was working perfectly, one should be able to derive the PH info from the lower level EventLog info -- we have not tested this, it would be a complex task.

Can you verify that my answer is correct?

3. In the EventLog table, what is the function/meaning of entries with action type InputResponse? For InputResponse entries the UserINput field is blank and the ActivityName field can take on a large range of values. How do we interpret these ActitivyName values (for InputResponse)?

4. In the EventLog, what is the difference between elapsedTime and probElapsed? • Why do they increase steadily (across multiple problems) until a certain point in time and then go back to 0 (in the middle of a problem)? • Why are there zeros inserted into the middle of the increasing sequence for probElapsed eventlog-juneJul2017.xlsx stuProbHistory_12_argent_classes_summer_2019.xlsx

   • also, what is the difference between time and clickTime?

I am attaching some data files that we are looking at that spawn these questions.

Additional Q's from Yvonne:

Is the solution hint counted in numHints and numHintsBeforeSolve?

Is the solution hint counted as one of the first three hints (for timeToFirstHint, timeToSecondHint, and timeToThirdHint)?

What is happening when numHints > numHintsBeforeSolve? There are 9 cases of this in the sample data, and all of them have numMistakes = 0, numAttemptsToSolve = 1, and isSolved = 1. For 7 of these cases, effort == SHINT and for the remaining 2, effort == “unknown”.

What causes effort to be “unknown”?

Some students have very high numbers of numMistakes (i.e., 71) and numHints (i.e., 35). Are these variables just counting up the total number of mistakes/hints, even if the student makes the same mistake or views the same hint over and over and over again?

[perspegrity5]

From Frank email 6/16/20

determine that the first 5 mentioned are milliseconds
• timeInSession timeInTutor , timeToFirstAttempt, timeToFirstHint, and timeToSolve

For elapsed time and probElapsed they are not divided by 1000 anywhere in the code but they seem to be interacting with other data that is surely in milliseconds.

clickTime comes from the db as a timestamp type.

time is milliseconds since some date like 1970

[IvonArroyo]

Hi all, I believe these questions are solved given my last email with tom, where most of these were cleared up. --Ivon

[perspegrity5]

New gsheet up showing units is HERE.

Below are the remaining questions for this item. Ivon can you answer?

What is happening when numHints > numHintsBeforeSolve? There are 9 cases of this in the sample data, and all of them have numMistakes = 0, numAttemptsToSolve = 1, and isSolved = 1. For 7 of these cases, effort == SHINT and for the remaining 2, effort == “unknown”.

Is the solution hint counted in numHints and numHintsBeforeSolve?

Is the solution hint counted as one of the first three hints (for timeToFirstHint, timeToSecondHint, and timeToThirdHint)?

What causes effort to be “unknown”?

Some students have very high numbers of numMistakes (i.e., 71) and numHints (i.e., 35). Are these variables just counting up the total number of mistakes/hints, even if the student makes the same mistake or views the same hint over and over and over again?