How well do people remember tweets and news headlines when they offer inconsistent messaging? A series of two studies published in Applied Cognitive Psychology looked into this question, finding that participants had better memory for tweets compared to news headlines, irrespective of the messaging (in)consistency.
Today, people can access information through a variety of online sources. How does our brain make sense of a news headline communicating one message – such as, standardizing testing is a fair assessment tool – and a tweet that contradicts it – standardized testing is a limited tool. In this work, Tori Peña and colleagues examined how message inconsistency affects memory.
Seventy-two Stony Brook undergraduate students who were fluent in English participated in Study 1, which was conducted in an in-person laboratory. Information consistency was manipulated between subjects, with participants randomly assigned to either the consistent or inconsistent condition.
All participants saw both tweets and news headlines presented in randomized order, for a total of 192 items (96 headlines and 96 tweets). Topics included “Standardized Testing, Bitcoin, Brain Games, Keto Diet, Minimum Wage Raise, [and] Plastic Straw Ban.” All items were from real accounts or news media outlets. The position (i.e., pro/anti) of the type of information (i.e., news headline or tweet) was counterbalanced across the study items.
Each item was presented on screen for 15 seconds and the type of information was labeled above the statement itself (i.e., “News Headline” or “Tweet”). Participants rated items for self-relevance on a scale of 1 (not at all) to 5 (highly) which served as a way to maintain participants’ attention for each item. After this task, participants were prompted to recall as many of the statements as they could.
Next, they played an unrelated game (i.e., virtual Solitaire) for approximately 50 minutes, which served as a delay period. Lastly, participants saw 96 items, 48 of which were items they had already seen. They rated each item as “Old” or “New” and provided a confidence rating on a scale of 1 (not at all confident) to 3 (very confident). Lastly, they responded to exploratory questions secondary to the research question of interest. Study 1 lasted two hours.
Study 2 was conducted online but largely followed the same procedure. A total of 72 undergraduate students from Stony Brook University were included. In this case, participants played Sudoku for 3 minutes prior to proceeding to the free recall task. After the recall task, they played Sudoku for another 10 minutes prior to completing the recognition task. Study 2 lasted one hour.
Peña and colleagues found better memory for tweets over news headlines in the case of both consistent and inconsistent messaging across the types of sources. This was the case for both experiments (i.e., in-person and virtual) and both types of memory measures (i.e., free recall and recognition). Both studies also revealed that tweets elicited higher recall, more accurate recognition memory and greater confidence, compared to news headlines.
The authors did not directly measure perceived item credibility. They write, “It remains a possibility that participants did not perceive news headlines as more credible than tweets during the study phase even though, in general, people consider official news sources as more credible than social media, and report that they trust social media sources less than news sources.”
The research, “Memory for tweets versus headlines: Does message consistency matter?” was authored by Tori Peña, Raeya Maswood, Melissa Chen, and Suparna Rajaram.