Primary Outcomes (explanation)
Subsequent placement beliefs about saving behavior, and self and other beliefs
This will comprise the difference between other-beliefs and self-beliefs. Self-beliefs and other-beliefs will each be measured using two corresponding items (e.g. ‘How good is your household in saving?’ and ‘How good are similar households in saving?’ on an 11-point scale ranging from 0, not at all good – 10, very good). We will analyse the difference between the corresponding self and other items separately, and if Cronbachs alpha is at least 0.6 for the self-beliefs and for the other-beliefs, we will also make one construct for self-beliefs and other beliefs and calculate the difference between those two.
Subsequent placement beliefs = self-beliefs – other-beliefs
We will explore the effects of the information treatment on self and other-beliefs separately. For this, we will pool the data of the current study with the data from our first study (AEARCTR-0015428).
Theory of planned behavior constructs
Saving attitudes, subjective norms, perceived behavioral control and intentions are measured following standard practice using the theory of planned behavior (e.g. Fishbein & Ajzen 2010; https://people.umass.edu/aizen/pdf/tpb.measurement.pdf). We define the saving behavior in these construct as ‘saving money every month for the next 6 months’. And in each question we explained what was meant by this ('By saving in a month, we mean that you do not spend all of that month's income, and that you thus have money left over that month').
Attitudes
Attitudes are measured using six items on 7-point bi-polar scales asking about how respondents’ would feel if they were to save money every month for the next 6 months (e.g. ‘1, bad – 7, good’, ‘1, unnecessary – 7, necessary’). We will combine these into one construct if Cronbachs alpha is at least 0.6.
Subjective norms
Subjective norms are measured with two questions. The first question measures injunctive norms using three items on 7 point-scales ranging from '1, totally disagree' to '7, totally agree' (e.g. ‘Most people who are important to me would approve if I saved money every month for the next 6 months.’). We will combine these into one construct if Cronbachs alpha is at least 0.6.
The second question measures descriptive norms (‘Most people like me will save money every month in the next 6 months.’) on a 7-point scale ranging from '1, very unlikely' to '7, very likely'.
We will combine the constructs for injunctive and descriptive norms into one construct for subjective norms if Cronbachs alpha is at least 0.6.
Perceived behavioral control
Perceived behavioral control is measured using four items on 7 point-scales ranging from '1, totally disagree' to '7, totally agree' (e.g. ‘I am confident that I can save money every month for the next 6 months, if I wanted to’ and ‘It is entirely up to me whether I will save money every month in the next 6 months, if I wanted to.’). We will combine these into one construct if Cronbachs alpha is at least 0.6.
Intentions
Saving intentions will be measured using three items on 7 point-scales ranging from '1, totally disagree' to '7, totally agree' (e.g. ‘I intent to save money every month for the next 6 months’ and ‘I expect to save money every month for the next 6 months’). We will combine these into one construct if Cronbachs alpha is at least 0.6.
As a robustness check, we also measure intentions by asking in how many months of the next 6 months respondents intent to save money, with answer options ranging from 0 months till 6 months.