Table 6.

Summary of the results from ‘ data collection ’ studies included in the review (n = 26 studies)

Barriers	Strategies
Language or literacy problems	• Measures in other languages or bilingual interviewers [45,63,85,90,96,100,111,113].
	• Use of multi-media [48,106], or computer data collection [48,59,120].
	• Avoid self-administered surveys [117] use of objective data [97,117].
	• Short surveys [106].
Lack of landline telephone (for population based telephone surveys) or highly mobile population	• Mailed survey instead of telephone [90]*.
	• Supplement telephone with face-to-face surveys [49,110].
	• Online surveying [52,54,68].
	• Use of objective (e.g. tobacco sales) data instead of self-report [97,117].
	• Flexible data collection [60,113,114].
Mistrust of researchers and the use of the data	• Culturally trained interviewers [56,111] or locals [55,74,96,98,105,115,118,124].
	• Need to pilot test measures [100,114,115].

*Indicates good evidence from randomised controlled trial (see also Table 1).