Small sample sizes reduce precision of estimation. Try to acquire more data but avoid simply adding “more of the same” as obtaining lots of small samples from the same population will not help. Applied more broadly: seek different opinions to avoid confirmation bias.