|
Field
Intervention (Hidden)
|
Before
Subjects will be randomized into one of four experimental conditions. (The probability for any subject of being randomized into conditions 1, 2, or 4 is 24%, and probability for condition 3 is 26%). For those subjects assigned to conditions 1-3, the probability of receiving the ICT question with vs. without the sensitive item is 50%. The experimental conditions are as follows (all are translated by the author from the Spanish):
1. Direct questioning. The respondent is asked: “Did you report a greater number of heads than you actually obtained?”. If the answer is “yes,” the respondent is then asked: “How many heads did you really obtain?”.
2. Indirect questioning. The respondent is asked: “Do you believe that some participant similar to you reported more heads than they really obtained?”. If the answer is “yes,” then the respondent is asked: “How many heads do you believe that participant really obtained?”.
3. Randomized response technique. The respondent is asked to think of a friend whose day of birth they know. The respondent is shown two questions, and told to answer the first one if the day is between the 1st and the 7th of the month, and the second question otherwise. The first question is: “Are you an ITAM student?” The second question is: “Did you report more heads than you actually obtained?” The answer options are Yes or No.
4. Item count technique (or list experiment). Half of the respondents are shown a list of four items and asked to say how many are true (but not which ones): “I study full time and also work besides studying,”This may sound like an oxymoron but it is not uncommon in Mexico. “My BA thesis is (or will be) about monetary policy,” “I believe that decisions about economic policy often obey political considerations,” and “I voted in the elections in June of this year.” The other half of the sample are shown the same list plus an additional item: “I reported more heads than I really obtained.”
Note: On indirect questioning see Svensson (2002); on the RRT see Gingerich (2010), on the ICT see Glynn (2013).
|
After
Subjects will be randomized into one of four experimental conditions. (The probability for any subject of being randomized into conditions 1, 2, or 4 is 24%, and probability for condition 3 is 26%). For those subjects assigned to conditions 1-3, the probability of receiving the ICT question with vs. without the sensitive item is 50%. The experimental conditions are as follows (all are translated by the author from the Spanish):
1. Direct questioning. The respondent is asked: “Did you report a greater number of heads than you actually obtained?”. If the answer is “yes,” the respondent is then asked: “How many heads did you really obtain?”.
2. Indirect questioning. The respondent is asked: “Do you believe that some participant similar to you reported more heads than they really obtained?”. If the answer is “yes,” then the respondent is asked: “How many heads do you believe that participant really obtained?”. [Note, this was modified after the pilot study, August 14 2015, see updated pre-registration document attached]
3. Randomized response technique. The respondent is asked to think of a friend whose day of birth they know. The respondent is shown two questions, and told to answer the first one if the day is between the 1st and the 7th of the month, and the second question otherwise. The first question is: “Are you an ITAM student?” The second question is: “Did you report more heads than you actually obtained?” The answer options are Yes or No.
4. Item count technique (or list experiment). Half of the respondents are shown a list of four items and asked to say how many are true (but not which ones): “I study full time and also work besides studying,”This may sound like an oxymoron but it is not uncommon in Mexico. “My BA thesis is (or will be) about monetary policy,” “I believe that decisions about economic policy often obey political considerations,” and “I voted in the elections in June of this year.” The other half of the sample are shown the same list plus an additional item: “I reported more heads than I really obtained.”
Note: On indirect questioning see Svensson (2002); on the RRT see Gingerich (2010), on the ICT see Glynn (2013).
|