Primary Outcomes (explanation)
Overconfidence will be defined as the difference between beliefs about self and beliefs about others. For both the vague and specific condition, respondents will answer 3 items about their own saving behavior and 3 items about others’ saving behavior. We will calculate one overconfidence measure for the vague condition and one for the specific condition.
The overconfidence variable will be constructed such that a value larger than zero indicates overconfidence, and a value smaller than zero indicates underconfidence. A value of zero indicates no bias.
Since the vague and specific belief items will be measured on different scales, we will normalize the belief measures before we calculate the overconfidence measures, so that comparisons between the specific condition and the vague condition can be made. We will use min-max normalization using the scale end-points.
Below we describe in more detail how we will construct our overconfidence variable.
Vague Overconfidence measure
Before respondents answer the belief questions, they will receive information about how we define savings.
Explanation text about savings – The next questions are about saving behavior. By saving we mean money that is set aside and kept on hand. Savings can be held as cash, it can be money that remains in checking accounts, or money that is in savings accounts.
We will use the following belief items for the vague overconfidence measure.
Vague Beliefs about Self (VBS)
QUESTION TEXT VBS – We are curious what you think about your saving behavior.
Give your best estimate.
VBS1 – How good are you in saving money?
VBS2 – How good are you in building up savings?
VBS3 – How well do you handle your savings?
On an 11-point scale ranging from ‘0, not good at all’ to ‘10, very good’ / ‘0, not well at all’ to ’10, very well’
The order of the items will be counterbalanced.
Vague Beliefs about Others (VBO)
QUESTION TEXT VBO – We are curious what you think about the saving behavior of peers in the same life situation.
Give your best estimate.
VBO1 – How good are peers in the same life situation in saving money?
VBO2 – How good are peers in the same life situation in building up savings?
VBO3 – How well do peers in the same life situation handle their savings?
On an 11-point scale ranging from ‘0, not good at all’ to ‘10, very good’ / ‘0, not well at all’ to ’10, very well’
The order of the items will be counterbalanced.
Vague Overconfidence
Since all three belief-pairs are interpreted in the same direction, i.e. a higher score indicates better saving behavior, we can treat each item-pair the same way. Overconfidence for each belief pair will be calculated by subtracting the score for other beliefs from the score for personal beliefs for each item on the individual level. Each self and other belief measure will be normalized before we calculate overconfidence using 0 and 10 as the end-points of the scale for each belief item.
OverconfidenceV1,2,3i = VBS1,2,3i - VBO1,2,3i
Next, the three scores will be averaged into one overconfidence score, OverconfidenceVi, if Cronbach’s alpha is at least acceptable and 0.61 or higher. We will perform robustness tests on the separate OverconfidenceV1,2,3i measures.
Specific Overconfidence measure
Before respondents answer the belief questions, they will receive information about how we define savings.
Explanation text about savings – The next questions are about saving behavior. By saving we mean money that is set aside and kept on hand. Savings can be held as cash, it can be money that remains in checking accounts, or money that is in savings accounts.
We will use the following belief items for the specific overconfidence measure.
Specific Beliefs about Self (SBS)
QUESTION TEXT SBS – We are curious what you think about your saving behavior over the past 12 months.
Give your best estimate.
SBS1 – In how many months did you have savings? That is, you had more than €0 in savings.
SBS2 – In how many months did your savings increase? That is, at the end of the month, you had more savings than at the beginning of the month.
SBS3 – In how many months did your savings decrease? That is, at the end of the month, you had fewer savings than at the beginning of the month.
Answer field will be an open field. For item 1, the minimum answer that can be given is ‘0 months’ and the maximum is ‘12 months’, meaning there are 13 answer options. To prevent conflicting answers, the answer given at item 1, is the maximum that can be given at the next item, either 2 or 3, depending on the randomization of 2 and 3. Then, the maximum that can be given at the following item, again 2 or 3 depending on the randomization, is the value given at the first item, minus the value given at the second item that was presented.
Item 1 will be presented first, the order of items 2 and 3 will be counterbalanced.
Specific Beliefs about Others (SBO)
QUESTION TEXT SBO – We are curious what you think about the saving behavior of peers in the same life situation over the past 12 months.
Give your best estimate.
SBO1 – In how many months did peers in the same living situation have savings? That is, they had more than €0 in savings.
SBO2 – In how many months did the savings of peers in the same living situation increase? That is, at the end of the month, they had more savings than at the beginning of the month.
SBO3 – In how many months did the savings of peers in the same living situation decrease? That is, at the end of the month, they had fewer savings than at the beginning of the month.
Answer field will be an open field. For item 1, the minimum answer that can be given is ‘0 months’ and the maximum is ‘12 months’, meaning there are 13 answer options. To prevent conflicting answers, the answer given at item 1, is the maximum that can be given at the next item, either 2 or 3, depending on the randomization of 2 and 3. Then, the maximum that can be given at the following item, again 2 or 3 depending on the randomization, is the value given at the first item, minus the value given at the second item that was presented.
Item 1 will be presented first, the order of items 2 and 3 will be counterbalanced.
Specific Overconfidence
A higher score on specific self and other beliefs 1 and 2 indicates better saving behavior, but for self and other beliefs 3 the reverse is true. Therefore, items 3 will be reverse coded. Overconfidence for each belief pair will be calculated by subtracting the score for other beliefs from the score for personal beliefs on the individual level. Each self and other belief measure will be normalized before we calculate overconfidence, using 0 and 12 as the end-points of the scale for each belief item.
OverconfidenceS1,2,3i = SBS1,2,3i - SBO1,2,3i
Overconfidence for the specific condition is calculated by averaging the three overconfidence scores into one score, OverconfidenceSi, if Cronbach’s alpha is at least acceptable and 0.61 or higher. We will perform robustness tests on the separate OverconfidenceS1,2,3i measures.
Beliefs about self will be defined as the average of the three items respondents answered about their own saving behavior if Cronbach’s alpha is at least acceptable and 0.61 or higher. It will be calculated on the individual level of the respondent. For respondents in the vague condition, it will be the average of items VBS1, VBS2 and VBS3. For respondents in the specific condition it will be the average of items SBS1, SBS1 and SBS3 (SBS3 will be reverse coded). We will use the normalized scores for each belief item.
Beliefs about others will be defined as the average of the three items respondents answered about the saving behavior of others if Cronbach’s alpha is at least acceptable and 0.61 or higher. It will be calculated on the individual level of the respondent. For respondents in the vague condition, it will be the average of items VBO1, VBO2 and VBO3. For respondents in the specific condition it will be the average of items SBO1, SBO1 and SBO3 (SBO3 will be reverse coded). We will use the normalized scores for each belief item.