1) The vast majority of people – most likely including you – are uncomfortable about telling others about their real income. Respondents may evade questions or give inaccurate answers (for various reasons, such as lack of trust, taxation-related concerns, etc.). Before the data collector starts the interview, s/he needs to gain the respondent's trust – therefore, ensure that the data collectors explain carefully why your agency needs the data, how it will (not) be used and why it is important that the information the respondent provides is correct. You might also consider using intervals (e.g. less than 200 USD, 200 USD to 300 USD, etc.) instead of asking for an exact salary, which could also help.
2) An alternative and possibly much easier way of collecting income data is asking the employers to inform you about the salaries they have paid to those employees who benefited from your project’s support (unless this is against a local data protection law). This requirement can already be included in the agreement that specifies the conditions of the project’s support.
3) If you supported jobs that provide different incomes over time (as is the case for construction, tourism, agriculture, and many other sectors), ensure that the baseline and endline data is collected for the same period of year (e.g. always for the same 3 months). Otherwise, you might end up with two sets of data that is not comparable.
4) As much as possible, verify the stated income from additional sources, such as the employer’s records (payroll) or verbal confirmation.
5) If your aim is to ensure that the target group members receive a decent salary for their work, consider rephrasing the indicator to “% of workers with a monthly income of more than [specify the amount and currency]. The amount can be, for example, the minimum income, the average income in the given sector, etc.