My cousin was recently killed when hit by a car from behind as he was cycling his road bike in the countryside in broad daylight. The accident occurred on a sunny day as he entered a shaded area of tree's. I am now looking at raising awareness of the benefit of using rear lights in the day and I could use some help please.
No. Such studies are rarely done, and it would be difficult to get ethical approval for another one as the benefits have been clearly established. This would mean asking the people not using lights to put their lives at risk merely to reproduce other studies. Even if such approval could be obtained, I think the study should not be undertaken for that reason.
Edit: since people seem to be struggling to understand the question and my answer:
the question was specifically "recent studies", hence the "no" answer.
The ethics of the study might vary between social science and medicine/engineering, but in the latter you cannot generally ask or permit people to take a dangerous action, no matter how interesting it might be. A study comparing, say, constant with flashing lights is more likely to get ethical approval than one which asks people not to use lights, or just dumps them from the study if they do. Some social scientists (eg, economists) have less restrictive ethical rules.
the difference is between passively observing that many cyclists ride without daytime lights, and asking people to refrain from using lights during the day. The latter requires explicitly saying "this is dangerous, please do it", and that is where ethical problems arise.
Imagine the question is, say "is smoking dangerous", or "is being obese dangerous" now. You can't really ask people to take up smoking or become obese (alternatively, that request is unlikely to get ethical approval), and you have to carefully design the study to allow for people who stop being eligible during the study. But, importantly, when you're comparing a fix (a quit smoking program, or a diet change), you need to come up with a control intervention that reveals the (dis)benefit of the chosen intervention without further harming the participants. Normal practice is to compare the intervention under study to the current best practice intervention rather than to nothing. There is even a meta-study of how often trials are stopped early for benefit when it becomes obvious that the trial intervention is significantly better than the comparison.
So for "cyclists using lights during the day" questions, a comparison of best practice (constant daylight lights) to a plausible alternative (flashing lights) is more likely to be approved than one where the alternative is the known-dangerous, definitely not best practice "no lights".
The only study I have done is personal experience, I have been hit during broad daylight both with lights on and no lights.
Bikes are not difficult to see during the day, so when they're in an accident (if it's not the bikes fault) it's usually an inattentive driver (being kind here) and lights would make little to no difference. Or it's a vehicle trying to squeeze through a small gap and misjudging. My latest (last week) was a bus which hit me twice crowding me off the road and the drivers explanation when I tracked him down and talked to him was there wasn't enough room, not that he didn't see me.