Erratum

Since the publication of this article [1], we have been made aware of two errors in the manuscript.

First, it was reported that 344 participants were allocated to one of two groups. This was a typo consistent throughout the article, and should have been 334 (168 intervention, 166 control – which was reported correctly).

Secondly, the RR calculation was based on the transposed table, which produced the correct p value and OR, but incorrect RR (1.38 instead of 1.51). As such, it was reported that: “those sent the personalised email were 1.5 times (95 % CI = 1.18–1.93) more likely to respond than those sent the generic email.” This should have read: “those sent the personalised email were 1.4 times (95 % CI = 1.15–1.66) more likely to respond than those sent the generic email.”