Research shows that giving health advice in a commanding tone is more likely to be met with resistance

Health advice, such as abstaining from alcohol or smoking, often uses urgency-inducing language, such as 'You should stop now' or 'You must follow the advice.' However, a meta-analysis has shown that such strong language can create a sense of 'threat to freedom' in the recipient, triggering a defensive reaction and leading to resistance to the advice. The research team compiled multiple studies to examine the chain of events that lead to 'threat to freedom' language, such as commanding tone, losing persuasive power.

Words that trigger: a meta-analysis of threatening language, reaction, and persuasion in health | Journal of Communication | Oxford Academic

https://www.psypost.org/forceful-language-makes-people-resist-health-advice/

The study was conducted by Long Ma of Suzhou University of Science and Technology, Zhexin Ma of the University of Connecticut, and others, and was published in the academic journal Journal of Communication.

Ma and his team focused on ' psychological reactance theory.' People value the freedom and autonomy to choose their own actions, and when a message suggests that this freedom is taken away, a negative motivational state called 'reactance' is triggered. Psychological reactance theory explains that when reactance increases, recipients may ignore the message in an attempt to regain their autonomy, or in some cases, even engage in 'prohibited behavior.'

The purpose of this analysis was to clarify, across studies, how the 'phrasing' of health-related messages triggers defensive reactions and ultimately leads to the success or failure of persuasion. While similar phenomena have been examined in previous studies, the research team felt that there was a lack of understanding of how the factors that shape reactance are connected and which words are likely to trigger them.

The research team therefore conducted a meta-analysis, which combines data from multiple studies to estimate overall trends. They selected experimental studies that met their criteria from over 1,000 reports, ultimately selecting data from 35 studies covering a total of 10,658 people. All of the selected studies focused on individual health behaviors such as quitting smoking and drinking alcohol. The research team excluded studies on altruistic behaviors such as organ donation, as the psychological motivations may not be the same.

The research team examined the following chain of events:

Exposure to language that threatens freedom
↓
feel that one's freedom is threatened
↓
Increased anger and reactionary thoughts (state reactance)
↓
They are less likely to agree with the message and less likely to intend to change their behavior.

The research team coded the statistical results of the included studies to see if these connections held across studies.

The key point was how to handle 'state reactance.' Some positions see reactance primarily as 'anger,' while others see it as a function of thoughts such as 'forming a counterargument.' The research team adopted an 'entanglement model' that does not separate the two, but rather considers anger and negative thoughts to be intertwined and arise simultaneously.

The research team's analysis showed a statistical correlation between the use of language that threatens freedom and recipients' perception that 'their choices are restricted.' The research team stated, 'The more language implying a lack of choice, the more likely people are to feel threatened.' Furthermore, the perception of threat was strongly linked to state reactance, and people who felt their freedom was being restricted were more likely to feel angry and have more negative thoughts about the message.

The same analysis also found that people with higher state reactance were less likely to agree with the message and less likely to intend to change their behavior. However, the research team noted that while the negative association was consistent, the effect size was not large, suggesting that factors other than reactance may also be involved in health behavior decision-making.

The research team also looked at the expressions used to create the 'freedom-threatening phrases' in the experiments. As a result, the following expressions were combined in many of the studies:

・Expressions that strongly state what must be done, such as 'must' and 'have to'
- Categorical language that does not allow for exceptions
・Exclamation mark
・An expression that clearly states that there is no choice

On the other hand, messages perceived as a low threat to freedom tended to use phrases that conveyed advice rather than commands, such as 'could' and 'consider,' which left room for the recipient to choose whether to act or not.

However, the research team also noted that simply including more strong expressions does not necessarily result in a stronger reaction. Simply counting the number of elements that threaten freedom did not effectively predict the strength of the backlash. The research team suggested that the context and the combination of which words were used may be important.

The research team also examined whether the results varied depending on how state reactance was measured, but no statistically significant difference was found when comparing studies that measured only anger with studies that measured only negative thoughts. The research team states, 'The fact that resistance expressed as anger and resistance in the form of constructing counterarguments in one's head had roughly the same effect on persuasion is consistent with the entanglement model we adopted.'

However, the research team also raised some cautionary points. The results of the included studies varied widely, with the strength of the effect varying considerably from experiment to experiment. 'While we were able to confirm the overall trend, it is possible that our analysis does not fully explain the conditions under which the effect is stronger,' the research team stated. The research team also suggested that differences such as age and gender may have an impact, and noted that the data suggests that men tend to be more susceptible to persuasion being undermined by reactance than women.

Furthermore, the research team pointed out that the analysis only looked at textual expressions and did not include visual elements such as image stimuli or color schemes. Since visuals can be a source of resistance when it comes to health-related messages, future research should also consider how design influences the perception of 'threats to freedom.'

Regarding future challenges, the research team stated, 'It is necessary to isolate which expressions are the strongest triggers.' Past studies have often incorporated multiple elements simultaneously, such as imperative expressions, assertions, and exclamation marks, making it difficult to independently evaluate which element is most effective. The research team concluded that further research is needed to manipulate and compare elements one by one, to take into account differences in cultural backgrounds, and to examine whether there are differences depending on the type of health behavior.