A few months ago, Nielsen released a study that found an increase in Twitter volume about live TV correlated with an increase in TV ratings across varying age groups. Now Nielsen is out with another study, with a bit more of a definitive tone. Most notably, the term “correlation” from the previous research has been replaced by “causal influence” in this one. Per the researchers, “the volume of tweets caused significant changes in live TV ratings among 29 percent of the [221 broadcast primetime program] episodes” sampled.
To arrive at that conclusion, Nielsen says it analyzed “minute-to-minute trends in Nielsen’s live TV ratings and tweets for 221 broadcast primetime program episodes using Nielsen’s SocialGuide.”
It’s worth noting that a statistically significant causal influence was only found in 29% of the episodes studied. In other words, there was no statistically significant impact for 71% of the episodes sampled.
It’s also worth remembering that sampled episodes were all live broadcast primetime. That raises the question of what impact might be present (or not present) for other program types, and whether or not the study paid undue attention to episodes that over-indexed in social engagement – questions raised in AdAge’s coverage of the study. That may be the case, although the latest data from Nielsen’s own Twitter TV Rankings suggest that during primetime, broadcast programs don’t have much of a leg up (if any) on cable programs in terms of tweeting volume.
Interestingly, tweets had more of an influence on some program genres than others. Impact was greatest among competitive reality shows: in 44% of such episodes measured, tweets were found to cause TV ratings changes. The impact of tweets was also higher for comedy programs (37% of episodes measured), while being about average for sports program episodes (28%) and below-average for drama episodes (19%).
The researchers also concluded that increases in live TV ratings during an episode were more likely to cause increased tweeting than vice versa. That is, TV ratings had an impact on related tweets among 48% of the sampled episodes.