Perceived need for climate information

Understanding that human activities primarily cause climate change is fundamental to driving behavioral change. However, a significant gap exists regarding its anthropogenic nature, with public understanding often lagging behind the scientific consensus. This underscores that previous efforts to bridge this gap have not been effective, highlighting that the reception of information is a complex process influenced by individual perspectives, biases, and preconceptions. Therefore, an individual’s information behavior is crucial in the interrelationship between climate science and the public. This study investigated the factors explaining individuals’ perceived need for information to form a firm opinion about climate change. It explored the roles of information gap, climate anxiety, and environmental adversity in shaping these needs. The study’s analysis is based on the 2021 Climate Change Opinion Survey (N = 76,328) from Meta’s Data for Good Program. The survey was conducted on Facebook, inviting active users aged 18 and above to participate through their news feeds. It was administered across 31 countries. For the analysis, multinomial logistic regression was employed. The results showed that a high climate information gap and anxiety are positively associated with information needs. Environmental adversity is also positively correlated with climate information needs, but one’s awareness of climate change moderated the association.


Introduction
The temperature and climate of the Earth have altered dramatically over the last millions of years.However, current and expected future changes vary in that they are primarily brought about by human behavior (IPCC 2007).Climate change is mainly caused by human behaviors, such as using fossil fuels, which emit greenhouse gases.It thus may be significantly alleviated by changes in such behaviors.Climate change is most likely to cause higher temperatures, floods, drought, extreme storms, heat-related mortality, infectious disease epidemics, and declines in food yields and water quality, all of which will significantly impact human life (Gifford et al 2011).Therefore, persuading people to adopt more environmentally friendly behaviors is essential (Depoux et al 2017).
People's attitudes are critical in bringing about behavioral change, but the public attitude remains divisive regarding climate change's anthropogenic nature.Despite the overwhelming consensus among climate scientists that human behavior is the primary driver of sped-up global warming, a substantial segment of the American public expresses skepticism about human activity's role in the process (Egan and Mullin 2017).Surveys on American public views on climate and energy (e.g., Funk and Hefferon 2019) also provide similar results.While 49% of Americans believed that human activity significantly contributed to climate change, 30% believed that it only had some influence, and 20% thought human activity has little to no effect on climate change.
Other extensive global surveys, such as the one conducted by Meta and Yale (2022) involving more than 100,000 people from 107 different nations, also provide similar evidence.The poll showed that less than half (45.84%) of respondents believed human activity was mostly to blame for climate change.34.58% considered it to be caused equally by human activities and natural changes, and 12.39% saw natural changes as the primary reason for climate change.The remaining respondents considered that other factors are at play or that climate change is not real.
The compelling evidence presented thus far unveils a persistent rift between the public and climate science.This divide may be attributed to what Robson and Robinson (2013) point out regarding the intricate nature of human behavior in finding, utilizing, and communicating information.To address climate change, efforts have been made to make climate information more accessible.Yet, this vital knowledge can sometimes be met with ignorance, denial, and rejection (Case et al 2005).It is a reminder that information reception is an intricate process influenced by the interplay of individual perspectives and biases.
Curiously, amidst the vast expanse of climate research, scant attention has been paid to understanding individuals' climate information behavior.Instead, the existing body of research predominantly centers on assessing individuals' knowledge, beliefs, and concerns regarding climate change (Crawley et al 2022).This oversight is a scholarly shortfall, for in our information-rich epoch, comprehending the intricate dynamics of how individuals engage with and respond to information stands as a paramount scientific quest (Sharot and Sunstein 2020).
This also has practical implications.Gaining insight into the factors that drive individuals' need for climate information is important in developing effective communication strategies.This understanding can guide the improvement of climate message reception and the motivation of the audience to seek further information about climate change.Consequently, this can increase the likelihood of positively influencing individuals through climate change communication.
Against this backdrop, this study investigated the factors that explain individuals' perceived need for information to form a firm opinion about climate change.This study is grounded in psychology literature that provides explanations for the mechanisms underlying the motivation to seek information and close knowledge gaps (e.g., Loewenstein 1994).It also draws upon communication literature connecting climate anxiety to information-seeking (e.g., Yang andKahlor 2013, Kahlor et al 2020).The study further augments this theoretical base by elucidating the need for climate information within the overarching context of environmental adversities.
The study proceeds as follows.The relevant literature is divided into four sub-sections in the following section, with relevant hypotheses being proposed.The measures and analysis methods are then discussed, followed by the findings.Finally, the results are discussed, and a conclusion is provided.

Information gap
How people engage with information is called information behavior (Bates 2010).Human behavior concerning information needs, seeking, and use has received attention in this field.In doing so, extant literature has linked information behavior with curiosity and information gaps.
Although curiosity is not the main variable of interest in this study, the concept is important for understanding the cognitive processes that underpin information-seeking behavior.Litman (2005) defines curiosity as the 'desire to know, to see, or to experience that motivates exploratory behavior directed towards acquiring new information' (p.793).Previous studies (e.g., Loewenstein 1994, Van Dijk andZeelenberg 2007) suggest that curiosity is often triggered by the realization of an information gap, driving individuals to seek to close it.In this context, an information gap represents a situation where people lack specific information and are motivated to obtain it (Noordewier and van Dijk 2017).
However, curiosity can also be motivated by interest and a desire for discovery rather than solely by a lack of information (Litman 2005, Litman 2008, 2010).This broader perspective implies that information behavior can also be attributed to casual leisure activities, such as web browsing, driven by hedonistic demands for pleasure rather than specific information needs (Elsweiler et al 2011).In scientific contexts like climate change opinion formation, Loewenstein's (1994) proposition becomes more applicable, suggesting that the primary motivation for needing and seeking information is to address an individual's lack of understanding.Hence, the information gap serves as the variable of interest in this study, which is evaluated based on one's climate awareness.Higher levels of climate awareness indicate a smaller information gap, while conversely, lower levels of climate awareness indicate a larger information gap.
Since it is characterized by indeterminacies, where there is insufficient information about the probabilities of a given set of outcomes, climate change is frequently associated with uncertainty.These uncertainties are about the extent, distribution, and timing of the effects of climate change (Visschers 2018), which blur the degree of climate change's current and future impacts.The limitations of measurement tools and the need to comprehend intricate and multifaceted underlying processes, such as natural disasters and other physical, biological, and human systems, further add to the uncertainty.Thus, not all information on climate change, including its effects in the future, is available.This presents a circumstance of a high information gap.
People can only anticipate the magnitude of the future effects of climate change.Global surveys, such as the 2022 climate survey by Meta and Yale, reflect these uncertainties.When asked about how serious a threat climate change will be in the next 20 years, close to half (46.70%) of the respondents claimed that it is a serious threat.34.30% of the individuals weighed it to be a moderate threat, 9.55% did not know, and 9% suggested there is no threat at all.
Given these findings, it is reasonable to expect that the perceived need for information to form firm opinions regarding climate change, which many people believe to be an impending occurrence (Weber 2010), will be positively associated with individuals' information gaps.The following hypothesis is proposed to investigate the association: H1: A higher climate information gap leads to a greater need for information about climate change.
While seeming apparent, the conundrum with the analyzed association arises from previous studies indicating that many individuals may overstate their awareness of climate change due to social desirability biases (Yang and Kahlor 2013).Like climate awareness, information needs can also be viewed as a positive social behavior and a means of learning.As a result, this may possibly affect how information needs are conveyed, notwithstanding the information gap.The association, therefore, merits investigation.

Climate anxiety
Curiosity behaves similarly to other drives, such as hunger, which urges eating.
Based on this idea, Loewenstein (1994) proposed that a modest bit of information acts as a priming dose that significantly raises curiosity.Although consuming information is enjoyable, satiation sets in after repeated exposure, and the information works to curb further interest.Kang et al (2009) found this to be consistent.Their study found that when individuals had little information, they had the greatest desire to know.
From this perspective, considering the current state of the availability of information on climate change or any other issues, it is conceivable that people may lose interest and tune out.This may be particularly relevant for scientific issues such as climate change because individuals and groups cannot effectively process and respond to information (Shome and Marx 2009).Describing climate change and climate variability is not a straightforward transmission of information (Weber 2010).Therefore, to understand what may motivate people to seek information beyond an information gap, the literature on climate anxiety and information-seeking is consulted.
Climate change is becoming recognized as posing several perils to human health, including mental health (Clayton andManning 2018, Manning andClayton 2018).Extreme weather events are found to hurt people's mental health.For instance, Hurricane Katrina's impact on psychological distress in the affected areas was still palpable 12 years after it struck (Raker et al 2019).Studies have also shown that heat waves affect mental health (Hansen et al 2008).
Research has shown that psychological anguish and negative emotional responses can also result from perceptions of climate change, in addition to actual experiences that cause it.Climate anxiety is perhaps the most frequent term for summarising these indirect negative emotional responses (Clayton 2020, Clayton andKarazsia 2020), and it can involve a persistent, uncontrollable apprehension and worry about climate change (van Valkengoed et al 2023).In such a state, individuals are said to look for information as an adaptive reaction (Charpentier et al 2022).This is because it allows one to engage in a risk control mechanism (Witte 1994) since information provides a sense of control, reduces uncertainty, and facilitates action (Kobayashi and Hsu 2019, Sharot and Sunstein 2020, Kelly and Sharot 2021).
Along these lines, there is evidence suggesting a connection between emotional responses regarding the environment and the tendency to engage in climate information-seeking behaviors (Yang and Kahlor 2013, Ho et al 2014, Hmielowski et al 2019, Kahlor et al 2020, Williams 2021).Yang and Kahlor (2013) investigated the effect of positive and negative emotions on information-seeking and avoidance in light of climate change.Their study showed that whereas negative emotions influenced information seeking, positive ones drove information avoidance.Ho et al (2014) also showed that negative affect was positively associated with information-seeking intention about climate change.
Other studies (e.g., Hmielowski et al 2019, Kahlor et al 2020) examining factors that trigger intentions to seek information on risk-related environmental issues have reported similar results on the influence of negative emotions on information seeking.Furthermore, digital behavioral studies (e.g., Williams 2021) have also illustrated a positive association between online information-seeking for climate change and mental healthrelated terms such as anxiety.
This study expands on previous research by examining whether the positive association between information-seeking and climate anxiety extends to the relationship between information needs and climate anxiety.The following hypothesis is proposed to investigate this association: H2: The more anxious individuals are about climate change, the higher their climate information needs.While closely related, information-seeking may not fully represent people's information needs.This is because the interpretation of needs is often influenced by contextual factors, such as the level of task urgency (Savolainen 2017).This study's focus on how people perceive the need for climate information to form a firm opinion aligns with this notion, recognizing people may feel they need information, but this may not always lead to them actually seeking it out or intending to do so.
An overarching theoretical framework that encompasses and explains the various indicators of climate anxiety has not yet been produced, and conceptualizations of climate anxiety vary significantly between studies.But apprehension and worry are the core characteristics of climate anxiety (van Valkengoed et al 2023), as in the case of anxiety in general (Amstadter 2008).Therefore, climate anxiety in this study is operationalized using worry about climate change.Environmental conditions influence people's emotional responses to climate change.

Environmental adversity
Given that literature also suggests that such emotional responses lead to information seeking (Yang andKahlor 2013, Ho et al 2014), it is reasonable to consider that people could need and seek information about climate change when exposed to adverse environmental conditions because they would be more concerned.To explore this, the following hypothesis is put forth for investigation: H3: Exposure to more adverse environmental conditions will increase the need for information about climate change.
However, the literature on this subject is equivocal, with studies (e.g., Whitmarsh 2008) suggesting that environmental conditions may have little or no effect on affective perceptions.Therefore, it stands to reason that the proposed information behavior could likewise change.Considering studies (e.g., Corner 2012, Shi et al 2016) imply that people's awareness of climate change affects their concern about it, the conflicting data regarding the effects of external factors on affective perceptions about climate change may be attributable to individual variances in climate awareness.Graham et al (2022) suggest that when people with a higher awareness of climate change are exposed to air pollution, they tend to be more concerned about climate change and its impacts on their health.Climate awareness brings an understanding of the negative effects of climate change and allows for identifying connections between climate change and health impacts (Yang et al 2018).Therefore, the concern about climate change is said to be preceded by knowledge of its consequences (Arvai et al 2012, Walsh andTsurusaki 2014).
One's level of climate awareness determines the relationship between exposure to unfavorable climaterelated situations and the degree of the ensuing emotional response.More severe adversity in the environmental condition combined with higher climate awareness may yield greater worry.Therefore, it is plausible to consider that more adverse environmental conditions accompanied by greater climate awareness will increase the need for climate information.
In other words, the influence of adverse environmental conditions on peoples' need for climate information is moderated by climate awareness.The following hypothesis is put up to investigate the association between environmental adversity, information needs, and climate awareness: H4: Climate awareness influences people's information needs in response to adverse environmental conditions.
In this study, environmental adversity is assessed based on national air pollution levels.Air pollution is employed because instead of seeing climate change as a standalone problem, the public views climate change as a component of a larger range of social and environmental challenges (Whitmarsh and Capstick 2018), such as air pollution (Darier andSchüle 1999, Wolf andMoser 2011).Whitmarsh (2008) suggests that many people may conceptually and experientially associate air pollution with climate change.Their study illustrated that the perception that climate change is occurring and being brought on by human activity correlates with one's experience with air pollution.
The public also frequently sees air pollution as a key driver of global climate change (Bostrom et al 1994).According to a study conducted in 10 Sub-Saharan nations, it was found that many Ethiopians claimed climate change to be caused by pollution or smoke resulting from local activities like burning firewood for fuel (Burton et al 2009).
Moreover, direct indicators of climate change, such as annual changes in temperature, may prove less conspicuous.Air pollution is one of the largest risks to public health globally that is deeply connected to climate change (Pinho-Gomes et al 2023), and numerous studies (e.g., Cisneros et al 2017, Pu et al 2019, Cori et al 2020) have demonstrated a correlation between the air quality of a particular location, as measured by PM2.5 levels, and the perceived risk associated with it.
It not only causes cardiopulmonary disorders and impairments, but PM2.5 has also been found to contribute to other detrimental health effects.These include promoting the onset and advancement of diabetes mellitus and influencing adverse birth outcomes (Feng et al 2016).Furthermore, studies have associated high PM2.5 levels with increased mortality rates for heart diseases and stroke (Al-Hemoud et al 2019).

Potential confounders
The study took into account seven covariates.Besides the participants' background characteristics, including age, gender, and education, the study included covariates for climate belief of whether or not climate change is occurring, perceived risk of climate change to oneself, climate institutional context, and urbanicity.
The acceptance or rejection of climate change is a crucial determinant of any stance or conduct on climate change.Therefore, it is essential to consider these beliefs regarding the occurrence of climate change in the analysis.In addition, a crucial element in climate change research is comprehending the associated level of risk (Arbuckle Jr et al 2015).Environmental behavior theories emphasize the role of risk perception, indicating that people's perception of their vulnerability to climate change hazards can influence their attitudes and actions.Heightening awareness of these risks can instill a greater sense of obligation and urgency in individuals to safeguard the environment, potentially resulting in increased pro-environmental behavior (Lieske et al 2014).Therefore, it is reasonable to consider risk perception may also influence the need for climate information.
The institutional context relating to climate change or how a country deals with climate change at an institutional level may influence the attitudes and actions of its citizens towards the environment.This can be attributed to the cognitive processes in forming perceptions about the issue, wherein individuals take cues from the prevailing institutional context.Institutions can play a vital role in defining the problem and context of climate change (O'Riordan and Jordan 1999).While institutions have been defined variedly, Young et al (1999) claim that 'At the most general level, institutions are constellations of rules, decision-making procedures, and programs that define social practices, assign roles to the participants in such practices, and govern the interactions among the occupants of those roles.' The inclusion of urbanicity, which refers to the level of urban development in an area, serves as a covariate in order to account for the variations in environmental conditions within nations.Urban areas are often considered the main source of pollution.Besides, cities' inherent structure and fabric make them prone to overheating.Due to less vegetation, heightened human activity, heat-absorbing qualities of urban construction materials, and limited ventilation and heat-trapping caused by high and dense infrastructure (IPCC 2021), cities are warmer than surrounding rural areas (i.e., urban heat islands).The urban heat island effect exacerbates heatrelated impacts of climate change in cities, making extremely hot weather more severe and difficult to manage.

Data
The study's analysis is based on the 2021 Climate Change Opinion Survey conducted by Meta (Facebook Data for Good initiative) in collaboration with the Yale Program on Climate Change Communication.The survey was fielded on Facebook between February 17, 2021, and March 3, 2021, and was administered in 31 countries.Users aged 18 and above were presented with an invitation to participate in a short survey at the top of their Facebook News Feed.Participation in the survey was voluntary, and all participants consented before proceeding.
The 31 countries include Argentina, Australia, Brazil, Canada, Colombia, Costa Rica, Czech Republic, Egypt, France, Germany, India, Indonesia, Ireland, Italy, Japan, Malaysia, Mexico, Netherlands, Nigeria, Philippines, Poland, Russia, Saudi Arabia, South Africa, Spain, Taiwan, Thailand, Turkey, United Kingdom, United States, Vietnam.The survey was limited to monthly active Facebook users.In order to enhance the data's representativeness to the demographics of each country, survey weights were applied (Meta and Yale 2021).

Participants and measures
The online survey attracted nearly 80,000 (N = 76,328) participants.The United States had the most participants (n = 13,555, 17.76%), followed by India (n = 4502, 5.90%) and Canada (n = 2888, 3.78%).Saudi Arabia had the fewest (n = 952, 1.25%).Regarding age, 16.28% of respondents were between the ages of 18 and 29, which was also the smallest group.The age group with the most respondents (32.50%) was above 60.Those aged 30-44 and 45-59 were comparable, with 23.95% and 27.25%, respectively.The majority of the respondents (53.25%) were female.Only 34.10% of the participants held a bachelor's degree or higher.
Refer to tables A1 and A2 in the Appendix for detailed descriptive statistics.
The survey employed an ordinal scale to measure climate attitudes.The reliability and validity of the scales used in the opinion survey are well-established, as evidenced by their extensive utilization in numerous reports and publications concerning public opinions and attitudes toward climate change (e.g., Leiserowitz et al 2021, Leiserowitz et al 2022).
Gender.The survey participants were asked, 'What is your gender?'The responses are coded as Female = 0 and Male = 1.
Education.The survey participants were asked, 'What is the highest level of education you have completed?'Due to the differences in education systems, the responses vary across countries and territories.Therefore, education is coded into two categories to simplify the melding of the various systems: less than a bachelor's degree = 0; bachelor's degree or higher = 1.

• Key variables
Information need.The survey participants were asked, 'On some issues, people feel that they have all the information they need in order to form a firm opinion, while on other issues, they would like more information before making up their mind.For climate change, where would you place yourself?'The responses are coded on a four-point scale: I do not need any more information = 1; I need a little more information = 2; I need some more information = 3; I need a lot more information = 4. 24.89% of respondents claimed they needed a lot more information, 34.74% needed some more, 18.32% needed a little more, and 22.04% didn't need any.
Climate awareness.Participants' awareness of climate change was assessed by asking, 'How much do you know about climate change?'The responses are coded on a four-point scale: I have never heard of it = 1; I know a little about it = 2; I know a moderate amount about it = 3; I know a lot about it = 4.The lowest percentage of respondents (4.41%) said they had never heard of climate change.A little and moderate knowledge of climate change was claimed by 30.7% and 47.8% of the participants, respectively.17.06% of those surveyed asserted knowing a lot about climate change.
Climate anxiety.The survey participants were asked, 'How worried are you about climate change?'The responses are coded on a four-point scale: Not at all worried = 1; Not very worried = 2; somewhat worried = 3; very worried = 4.Most survey respondents were either very worried (39.23%) or somewhat worried (38.3%) about climate change.The remaining 14.2% of respondents said they were not very worried, and 8.18% said they were not worried at all.
Environmental adversity.The annual average exposure of a population to concentrations of suspended particles (PM2.5), which potentially harm human health, is utilized in the study to indicate environmental adversity.
The data used in this study is the 2019 PM2.5 air pollution data acquired from the World Bank Open Data, which represents the most recent data available at the time of conducting this study.Based on this data, the countries were categorized into six groups according to their mean annual exposure levels.For countries with 1-10 mean annual exposure (micrograms per cubic meter) = 1; 11-20 = 2; 21-30 = 3; 31-40 = 4; 41-50 = 5; 50+ = 6.Most of the participants (40.56%) fell under group 2. The nations with the highest exposure accounted for 10.71% of the respondents, while the ones with the lowest exposure accounted for 34.2%.

• Covariates
Climate belief.The survey participants were asked, 'Climate change refers to the idea that the world's average temperature has been increasing over the past 150 years, will increase more in the future, and that the world's climate will change as a result.What do you think: Do you think that climate change is happening?'The responses are binary coded: Yes = 1 and No = 0. Most of the participants (93.77% ) believed climate change was happening.
Perceived risk.The survey participants were asked, 'How much do you think climate change will harm you personally?' The responses are coded on a four-point scale: Not at all = 1; only a little = 2; A moderate amount = 3; a great deal = 4.Most respondents said there was a moderate risk from climate change (35.6%) and a high risk (32.30%).20.38% considered it to be of little risk, and 11.68% thought it to be of no risk.
Institutional context.The number of climate laws and policies in a country is used to assess the institutional context of climate change.The data was sourced from the Grantham Research Institute's Climate Change Laws of the World database.The countries have been categorized into six groups based on the number of climate laws and policies they have implemented.These groups are denoted by the following codes: 1-10 policies = 1; 11-20 = 2; 21-30 = 3; 31-40 = 4; 41-50 = 5; over 50 policies = 6.Group 5 had the highest percentage of countries (31.42%), followed by Group 3 (26.48%).The least represented group, Group 2, contained 5.46% of the countries.
Urbanicity.The survey participants were asked, 'Which phrase best describes the area where you live?' The responses are coded on a four-point scale: Village or countryside =1; Town = 2; Suburb/Outskirts of a city = 3; City = 4.Most participants, equalling 40.28%, were from cities, and 17.87% were from the city outskirts.Participants from villages and towns made up 21.19% and 20.63%.
See figure A1 in the Appendix for the correlation between the study variables.

Analysis
Given the dependent variable's ordered nature in the present study, ordinal logistic regression was the preferred analysis method.The ordinal model, however, requires the data to meet the proportional odds assumption, which is a restrictive assumption that is often violated in practice (Fullerton and Xu 2012).Based on the Brant-Wald test, it was concluded that this study's model also did not satisfy the assumption.Therefore, a multinomial logistic regression was implemented to test the study's hypotheses.Since the study's data is stratified, meaning people are nested into countries, a mixed (multilevel) multinomial model was considered that allows the intercept to vary for each country.The random model was fitted using the 'mclogit' package in R 4.2.2.The sizes of the Odds Ratios (ORs) obtained from the analysis have been interpreted using the R package 'effectsize.'The interpretation is based on Chen et al (2010).They suggest that ORs that fall below 1.68 (or above 0.60) are to be considered very small, and those ranging from 1.68 to 3.47 (or 0.29 to 0.60) as small.ORs that range from 3.47 to 6.71 (or 0.15 to 0.29) and those that exceed 6.71 (or are less than 0.15) are to be classified as medium and large, respectively.
To indicate how much variance is explained by the random effect, the intraclass-correlation coefficient (ICC) was evaluated.The ICC was computed using the formula proposed by Bosker and Snijders (2011).In formula (1), t 00 is the random intercept variance or the between-subject variance.

Results
A mixed multinomial logistic regression was performed to create a model of the relationship between the predictors and the dependent variable in four groups of information need (i.e., 'no information,' 'little more information,' 'some more information,' and 'lot more information').The base category (i.e., no information needed) is the reference in the analysis.A model fit test was implemented utilizing the likelihood ratio test to compare the fit of the null and full multinomial models.The results revealed a significant difference in model fit (χ 2 = 19431.85,df = 30, p < .001),suggesting that the full multinomial model offered a significantly better fit to the data compared to the null model with only an intercept term.Additionally, the full model exhibited a much lower Akaike Information Criterion (AIC) value (AIC = 134014.40)compared to the null model (AIC = 153386.30).The lower AIC value for the full model suggests that it provides a more parsimonious representation of the data, balancing goodness of fit and model complexity.Furthermore, the Bayesian Information Criterion (BIC) also supports the fit of the full model, as it yielded a significantly lower BIC value for the full model (BIC = 134308.30)compared to the null model (BIC = 153413.10).
To further assess the goodness of fit, a Hosmer-Lemeshow test was performed on the full multinomial model.The Hosmer-Lemeshow test yielded a nonsignificant result (p = 1), indicating that the observed and expected frequencies in different groups were not significantly different.This suggests that the full multinomial model best fits the data, as it adequately predicts the observed outcome.
A fully unconditional model was fitted with the random component and no fixed factors using a cumulative logit link function to calculate the overall model's random effect covariance.The analysis resulted in a random intercept estimate of 0.42 (P < 0.001, 95% CI [0.26, 0.71]), showing the between-group component is significantly different from zero.Based on this, ICC (11.44) was calculated for the countries.This suggests that 11.44% of the response variation can be attributed to unknown factors related to countries.
Tables 1 and 2 provide the results of the mixed multinomial regression, showing the parameter estimates (standard errors) in table 1 and the ORs [95% CI] in table 2. Both the tables present the results in two models (i.e., without (A) and with (B) interaction terms).The effect sizes are provided in table 3.
H1 predicted that a higher climate information gap leads to a greater need for information about climate change.The findings showed that, when all other factors are held constant, climate awareness has a significant The variance-covariance matrix's diagonal elements hold the variances resulting from the random effect for the outcome categories, while the off-diagonal components hold the covariances resulting from the random effect between any two possible outcome categories.Significance: *** = p < 0.001; ** = p < 0.01; * = p < 0.05; Reference category = no information needed.
(p < 0.001) and negative relationship with information needs for each of the three information need categories: 'little more,' 'some more,' and 'lot more' in comparison to the reference.Thus, H1 is supported.Compared to the reference group, the likelihood that an individual will require a little more climate information dropped by 31% (OR = 0.69, 95% CI [0.67, 0.72]) for a unit increase in climate awareness.Similarly, a unit increase in climate awareness reduced the odds of one needing some more and a lot more climate information by 45% (OR = 0.55, 95% CI [0.53, 0.57]) and 54% (OR = 0.46, 95% CI [0.45, 0.48]), respectively.H2 predicted that the more anxious individuals are about climate change, the higher their climate information needs will be.The findings showed that given all of the other variables in the model are held constant, climate anxiety has a significant positive correlation (p < 0.001) with the need for climate information across all three categories: 'little more,' 'some more,' and 'lot more,' relative to the reference.Thus, H2 is supported.For a unit increase in climate anxiety level, the odds of one needing a little more information increased by 32% (OR = 1.32, 95% CI [1.27,1.38]).The likelihood of one needing some more and a lot more climate information was even higher for a unit increase in climate anxiety level, equalling 46% (OR = 1.46, 95% CI [1.40, 1.51]) and 56% (OR = 1.56, 95% CI [1.49, 1.63]) relative to the reference group.
H3 predicted that exposure to more adverse environmental conditions would increase the need for information about climate change.The results revealed that when controlling for all other variables in the model, the association between environmental adversity and the need for little climate information, compared to the reference category, was not statistically significant.However, a significant and positive association was observed between environmental adversity and the need for some more (p < 0.05) and a lot more (p < 0.001) climate information compared to the reference group.These findings provide evidence in support of Hypothesis 3. Thus, H3 is supported.In comparison to the reference group, a unit increase in environmental adversity increases the likelihood of one needing some more and a lot more climate information by 16% (OR = 1.50, 95% CI [1.02, 1.33]) and 50% (OR = 1.50, 95% CI [1.23, 1.83]), respectively.H4 predicted climate awareness influences individuals' information needs in response to adverse environmental conditions.The results showed that when all other variables are held constant, climate awareness and environmental adversity interaction are positive and significant (p < 0.001) for all the three information need categories relative to the reference.Keeping all other variables in the model constant, with a unit increase in environmental adversity that is accompanied by a unit increase in climate awareness, the odds of an individual needing a little more, some more, and a lot more climate information relative to the reference increased by 10% (OR = 1.10, 95% CI [1.07, 1.13]), 18% (OR = 1.18, 95% CI [1.15,1.20]),and 20% (OR = 1.20, 95% CI [1.17, 1.23]), respectively.
From table 3, the OR effect sizes in this study varied between very small and medium.Climate awareness was the only key variable with a small effect, while all the other key factors had very small effects.Climate belief had a small to medium effect, and perceived risk had very small to small effects among the covariates.All the other covariates exhibited very small effects.

Unpacking the interaction
In light of the significant interactions in the regression results, it is imperative to perform further visual examinations to gain a better understanding of the interactions for each outcome category.Figure 1 shows how environmental adversity interacts with climate awareness to influence the need for climate information.
Based on figure 1, the interactions across the four levels of information needs: 'no information' (level 1), 'little more information' (level 2), 'some more information' (level 3), and 'lot more information' (level 4), are most evident for levels 1, 3, and 4.
In level 1 (no information), the likelihood that an individual may not need climate information is higher when they have higher degrees of climate awareness and are exposed to low environmental adversity (i.e., low levels of air pollution).However, the probability of not needing information decreases with increasing adversity.For example, when awareness is highest, the probability of not needing information reduces from roughly 45% to less than 5% when the adversity level rises from lowest to highest.Conversely, the probability of not needing information increases for high adversity levels but lower climate awareness.When the awareness level is at its lowest, increasing adversity increases the likelihood of not needing information from approximately 10% to 16%.
In level 3 (some more information), increasing environmental adversity increases the likelihood of needing information only for those with climate awareness in the highest two categories (i.e., those who claimed to have moderate and a lot about climate change).An increase in environmental adversity increases the probability of needing information from approximately 32% to 35% and 25% to 40% for the highest two categories of awareness, respectively.For those who claim to know nothing about climate change or only a little, the likelihood of needing information decreases with increasing adversity levels.
In level 4 (lot more information), an increase in environmental adversity leads to an increased need for information across all awareness levels.The distinction, however, lies in the percentage probability of needing climate information.The likelihood tends to decrease with decreasing climate awareness levels.For instance, for the highest level of climate awareness, increasing the level of environmental adversity from lowest to highest improves the likelihood of needing a lot of climate information from about 10% to 48%.The rise is relatively low, from 32% to 48%, for the lowest awareness level.

Discussion
This study investigated the factors explaining individuals' perceived information needs to form a firm opinion about climate change.
Firstly, whether a higher information gap leads to a greater need for climate information was examined.The study findings showed that people's information gap positively relates to their information needs.Individuals with a low information gap regarding climate change had lower information needs, while those with a higher information gap had greater information needs.This finding aligns with previous research indicating that a key motive for acquiring information is to come out of the state of information deprivation (Loewenstein 1994, Noordewier andvan Dijk 2017).Previous studies (e.g., Norton and Leaman 2004, Lorenzoni et al 2007) have suggested that individuals often misinterpret and even overstate their level of climate awareness, influenced by social desirability bias (Yang and Kahlor 2013).This led to considering whether such bias would also impact information needs, potentially leading to an overstatement of both awareness and the need for information, as both having awareness and needing information (learning) may be perceived as socially desirable behaviors.However, the findings of the study failed to provide any evidence supporting this notion.Secondly, even though consuming information can be enjoyable, people can become satiated and lose interest in the subject (e.g., Loewenstein 1994, Kang et al 2009) over time, and information gaps can become irrelevant.Also, opinion formation is a dynamic process characterized by progressive evolution and influenced by various factors (Chacoma and Zanette 2015).Therefore, identifying additional factors influencing an individual's need for climate information was considered.For this, firstly, whether climate anxiety is associated with information needs was investigated.
This notion stemmed from prior research, which demonstrated that individuals tend to search for information when they experience climate anxiety (e.g., Yang and Kahlor 2013, Kahlor et al 2020, Williams 2021).However, it remained uncertain whether this would extend to information needs.Informationseeking typically implies needs, but people interpret their needs differently depending on contextual circumstances like task urgency (Savolainen 2017).Thus, seeking may not adequately account for people's need for information.
The findings showed that the need for information about climate change was positively influenced by climate anxiety.The findings imply that anxious people perceive the need for and seek climate information as part of their risk-control approach (Witte 1994), for it helps them to feel in control, lowers ambiguity, and encourages action (Kobayashi and Hsu 2019, Sharot and Sunstein 2020, Kelly and Sharot 2021).
Following climate anxiety, this study investigated if exposure to more adverse environmental conditions would enhance the need for climate change information and whether one's climate awareness influenced the relationship.The findings revealed a clear trend: as individuals faced greater environmental adversity, their need for information about climate change correspondingly increased.Concurrently, the likelihood of individuals not needing climate information decreased.Furthermore, this relationship was found to be influenced and moderated by the level of climate awareness among individuals.
A rationale for the moderating role of climate awareness in the relationship between environmental adversity and information needs can be found in studies that claim that concern about climate change is preceded by awareness of its consequences (Arvai et al 2012, Walsh andTsurusaki 2014), such as identifying connections between climate change and health impacts (Yang et al 2018).The experience of adverse climate conditions knowing about the risk may raise concerns greater than the experience of adverse conditions without knowing much about the risk.In the same vein, Graham et al (2022) have demonstrated that when people with a higher awareness of climate change are exposed to air pollution, they tend to be more concerned about climate change and its impacts on their health.Based on this, it can be said that the need for climate information in the face of adverse environmental situations diverges due to different levels of climate awareness.
The study's key contribution lies in illuminating the interplay between environmental adversity, climate awareness, and information needs.While the primary literature on information behavior posits that a high information gap due to low awareness would lead to greater information needs among individuals, this research expands our understanding by demonstrating how contextual factors can influence this dynamic.In the face of environmental adversity, the moderating role of climate awareness is such that higher climate awareness corresponds to a greater need for information.
By challenging the prevailing assumptions of the information gap theory and emphasizing the contextual influences at play, this study advances our theoretical understanding of individuals' climate information behavior within the broader landscape of environmental challenges.

Practical suggestions
The findings of this study offer the following suggestions for practice.Previous studies (e.g., Norton and Leaman 2004, Lorenzoni et al 2007) suggest that people often misrepresent their knowledge of climate change by conflating general awareness with knowledge.As a result, although the public is generally aware of climate change, they do not thoroughly understand it.
In this context, communicators should know how the audience perceives their understanding of climate change.If people incorrectly believe they possess deeper knowledge about climate change because of their general awareness, they may not feel the need for further information.Therefore, communicators could take the initiative to articulate the extent to which current understanding falls short and correct any misinterpretations.This may create a sense of an information gap and motivate the audience to need more information about climate change.As a result, this could increase the chances of positively influencing people through communication about climate change.
Communicating climate change within the context of adverse environmental situations can potentially enhance individuals' need for climate information.However, it is important to recognize that the influence of such communication strategies may vary depending on individuals' level of climate awareness.

Limitations and future directions
As with all studies, this study also comes with some limitations.Future studies should consider addressing the following key limitations.
First, since this study relied on cross-sectional data, its results cannot be interpreted as causal.For instance, this study considered that climate anxiety drives one's need for information.However, previous studies also suggest that exposure to certain types of climate information evokes emotional responses, causing climate anxiety.These associations require further research.Second, climate perceptions in the study are measured based on single-item surveys and are a possible limitation.Third, factors such as attitudes toward science may influence climate information needs.Regrettably, this study could not include such variables in our analysis due to the unavailability of data.
Investigating alternative methods beyond articulating current information gaps and rectifying any misinterpretations to stimulate interest in climate-related information represents a promising avenue for further inquiry.
This study interpreted OR effect sizes based on Chen et al (2010).Their interpretation of the magnitudes of ORs is based on epidemiological data from a non-exposed population with an illness rate of 1%, which is a conservative approach and could be too stringent for current research.Despite this, the present study acknowledges that the small and very small effect sizes imply that the relationships investigated in this study require further confirmation through additional research.Still, the study's findings serve as a starting point for expanding the existing limited research on individuals' perceived information needs regarding climate change and forming opinions.

Conclusion
The study's findings showed that a higher climate information gap leads to a greater need for information about climate change.Emotional perceptions of individuals regarding climate change also positively correlated to their climate information needs.People's needs for climate information increased proportionately to how anxious they are about climate change.Last, the need for information about climate change increased as individuals were exposed to more adverse environmental conditions, but one's climate awareness moderated this relationship.Based on these findings, it can be said that people's information needs about climate change may be linked to their awareness and anxiety levels regarding climate change.Additionally, their environmental setting may impact their information needs, albeit this will depend on how aware they are of climate change.

ORCID iDs
Tenzin Tamang https:/ /orcid.org/0000-0002-5554-0608 One's emotional engagement with climate change depends on exposure to the ostensible effects of climate change (Reser et al 2014).Many infer climate change's reality by concluding their experiences with extreme weather (McDonald et al 2015).Harsh occurrences like flooding and hurricanes can make climate change seem more apparent and urgent, increasing unfavorable attitudes about it (Myers et al 2013, Bergquist et al 2019).

Figure 1 .
Figure 1.Interaction between environmental adversity and individual climate awareness predicting information need (percentage probability).

Table 1 .
Predictors of climate information needs: estimates (standard errors).