At the core of the investigation were two web-based randomized experiments conducted by Professor Asako Miura and her team at the Graduate School of Human Sciences. These experiments employed two distinct survey techniques to measure agreement with the provocative statement: “The mass media are harmful to society.” The first employed straightforward, direct questioning, a conventional survey method that asks participants explicitly whether they hold this view. In contrast, the second method utilized a list experiment—a technique designed to elicit more honest responses by indirectly assessing opinions through a set of non-sensitive items, thereby reducing social desirability bias and pressure.

The results were striking. When asked directly, 45.1% of the combined study participants agreed with the statement that mass media harm society. However, when the same sentiment was measured through the list experiment, agreement rates dropped substantially to 29.7%. This 15.4-percentage-point difference is statistically significant and consistent across both studies, indicating that traditional direct surveys may overstate negative attitudes. This overestimation suggests that respondents might be conforming to socially influenced norms that are critical of the media rather than solely conveying their private opinions.

Such findings have profound implications for understanding public opinion dynamics. In social contexts where media criticism is widely expressed and socially accepted, individuals may feel compelled to echo these sentiments openly, even if their private beliefs are less negative. This phenomenon challenges conventional wisdom in public opinion research that assumes direct questioning accurately captures authentic attitudes. Instead, it points to the complex interplay between individual beliefs and social environments that shape expressed views, particularly on contentious topics like media trustworthiness.

The research also highlights the importance of methodological rigor in opinion polling. The list experiment deployed in these studies operates by embedding the sensitive statement among a list of innocuous items and asking respondents how many items they agree with, rather than which specific items. This subtlety helps mitigate respondents’ inhibition about revealing unpopular or socially stigmatized opinions, thereby providing a more reliable estimate of true public sentiment. The divergence in results between the two methods underscores how survey design can markedly influence conclusion framing in social science research.

It is important to clarify that the study does not argue that media criticism is unwarranted or insincere. Rather, it emphasizes how survey methodologies can shape the appearance of public consensus, particularly when expressing critical views is socially reinforced. By revealing potential inflation in the reporting of negative media attitudes, the research encourages a more cautious interpretation of widespread media distrust portrayed in questionnaire-based polls. This nuanced perspective is critical in a media landscape where trust is fragile and public opinion impacts democratic processes and media accountability.

Moreover, the research acknowledges limitations related to sampling. Both surveys were conducted via nonprobability online samples rather than probability-based representative samples, which caution against generalizing the findings to the entire Japanese population. Nonetheless, the replicated pattern across two separate experimental designs strengthens the validity of the inference that social desirability effects play a role in the overreporting of negative media attitudes in Japan’s current digital environment.

Professor Asako Miura’s inquiry was motivated by a keen observation: the persistently high levels of media distrust shown in prior questionnaire surveys appeared unusually consistent, prompting questions about the influence of question framing on respondents’ answers. By pioneering the application of list experiments in this context, her research opens avenues for more refined public opinion assessment techniques that can better capture the complexities of socially charged attitudes.

The study’s broader scientific contribution lies in advancing methodological innovation in the social sciences, suggesting that indirect questioning methods should be more widely adopted to avoid skewed measurements of contentious public opinions. As digital communication reshapes social norms and public discourse, understanding how surveys reflect—or distort—true sentiments becomes essential for policymakers, media practitioners, and scholars alike who rely on data to gauge societal trends.

In sum, this research from the University of Osaka offers a revelatory examination of how attitudes toward the mass media in Japan may be misrepresented by traditional surveys. It sheds light on the subtle factors influencing opinion expression and highlights the critical role of experimental design in parsing the societal perception of institutions as influential as the mass media. The findings encourage a re-examination of how we gather and interpret public opinion data, particularly on issues that are heavily embedded in social discourse.

Subject of Research: People
Article Title: When Negative Media Attitudes Are Overreported: Comparing Direct and List-Experiment Measures in Japan
News Publication Date: 24-Jun-2026
Web References: http://dx.doi.org/10.1093/ijpor/edag013
References: Miura, A. (2026). When Negative Media Attitudes Are Overreported: Comparing Direct and List-Experiment Measures in Japan. International Journal of Public Opinion Research. DOI: 10.1093/ijpor/edag013
Image Credits: Asako Miura, International Journal of Public Opinion Research
Keywords: Social sciences, Social research, Anthropology, Mass media