I will briefly present archival and survey evidence to argue that we should be attending more to situations where measurement is not only imprecise but systematically skewed through the self-reporting process. Take participants’ age as an example: if research participants who are asked their age responded by rounding systematically to the nearest whole year, then means reported to 1-2 decimal places (the field’s norm) might technically be too precise, but they would at least be more or less accurate. Instead, what we see in practice is that the age mean computed from participant reports is approximately half a year lower than the age mean computed from their real dates of birth. As researchers handling self-reported age and similar types of data (e.g., length of relationships; duration of schooling) will naturally want to avoid this, I will conclude with practice recommendations.