In this study, the variable was operationalized using three dimensions already present in the original instrument:

Clarity refers to comprehensibility, precision, and consistency; channels represent accessibility and suitability of the media used; and feedback expresses the possibility of asking questions, obtaining answers, and recognizing that communication is bidirectional. These dimensions alone do not encompass all of strategic communication, so the results should be interpreted as a partial evaluation of the construct.

The social license to operate describes a continuously negotiated social acceptance. Thomson and Boutilier (2011) explain that it can progress from minimal acceptance to relationships based on trust and identification. Legitimacy is linked to the perception that activities are appropriate and respect shared norms; credibility requires congruence between words and actions; and trust implies a willingness to accept vulnerability to the behavior of another actor. In the Peruvian case, Saenz (2019) found that mining companies strengthen their social license when they articulate pragmatic, moral, and cognitive legitimacy with trust based on integrity and competence.

This framework allows us to avoid an instrumental interpretation according to which more intense communication should automatically produce environmental acceptance. A community can perfectly understand a message and, at the same time, disagree with the project. Therefore, a low correlation between communication and perception does not demonstrate the irrelevance of communication; This may indicate that different domains are being measured, that there is unobserved heterogeneity, or that the instrument does not capture dimensions such as trust, compliance, procedural fairness, or participation.

Risk perception refers to subjective judgments about the probability, severity, and acceptability of threats. Slovic (1987) showed that such judgments are not derived solely from technical information: familiarity, perceived control, fear, potential consequences, and trust all play a role. In socio-environmental conflicts, perception can intensify when benefits and risks are distributed unequally or when institutions are considered opaque. The present study did not directly measure perceived risk or institutional trust; these concepts are incorporated to theoretically define the phenomenon and guide subsequent research, not to causally explain the results obtained.

According to Córdova (2002) criteria, it is conceptualized as the way an individual detects the surrounding environment, leading them to interact with it. Additionally, Nagar (2006) adds that various factors inherent to the environment and the qualities of the individual are involved, making it a process of familiarizing oneself with the environment through sensory information.

Environmental perception encompasses how people recognize, interpret, and value their environment. The instrument analyzed:

Knowledge includes declared information or understanding; attitudes express valuation and a favorable disposition toward care; and perceived responsibility relates to the attribution of duties and protective actions. The total score summarizes these components, but does not equate to a direct measurement of mining risk, business confidence, or social license.

A quantitative, non-experimental, cross-sectional study with an analytical-correlational scope was conducted. Variables were observed without manipulation and at a single point in time. Inference is limited to statistical associations within the sample; the design does not establish temporal precedence or causality. The expansion to multiple regression aims to estimate adjusted associations between communication dimensions, not to demonstrate the effects of communication on environmental perception.

This research employed a basic type of investigation. According to Tamayo & Tamayo (2003), basic research, also known as pure or fundamental research, occurs within a theoretical framework with the main objective of expanding knowledge and proposing theories. It precisely uses techniques such as sampling, so that its results can be applied to more complex contexts than those directly studied. It does not focus on the practical application of its results, as it considers that task to belong to other actors, not the researcher, being fundamental for building theories and concepts that explain social phenomena, such as citizens’ environmental perception regarding mining projects.

The research approach of this article was quantitative. According to Hernández et al. (2014), the approach is employed as a set of sequential and systematic processes, where each phase must follow the previous one, without the possibility of skipping any or advancing. Although some phases may be adjusted, the order is maintained without modification. The process begins with an idea that is defined, objectives and questions are established, the literature is reviewed, and a theoretical framework is created. Then, hypotheses are formulated, their testing is planned, and data is collected and analyzed to draw conclusions.

Regarding the design, it was non-experimental. According to Hernández et al. (2014), in this type of design, no variable is manipulated. In other words, the independent variable is not altered to assess its impact on other variables. Additionally, it is cross-sectional. In this sense, Hernández et al. (2014) notes that data is collected at a single point in time, with the purpose of detailing and evaluating the impact and interdependence at a specific moment.

To determine if the research has correlation, it is important to distinguish between the two variables that determine the variation among the factors. It is worth emphasizing that the main characteristic of correlational research is the straightforward handling of multiple variables simultaneously. As stated by Tamayo (2012), it is important to keep in mind that this covariation does not imply that there are causal relationships between the values.

The fieldwork was conducted during 2025 in the urban area of the Cañaris district, Ferreñafe province, Lambayeque. At that time, the Cañariaco project was under the control of Alta Copper Corp. The acquisition by Fortescue was completed on March 10, 2026, after the data collection; therefore, the items were answered with reference to the project and the communication perceived within the corporate context of 2025.

The reference population consisted of 183 adult citizens linked to the area delimited for the study. The required sample size was calculated using the formula for proportions in finite populations:

n=(N×Z2×p×q)/[e2×(N−1)+Z2×p×q]

With N = 183, Z = 1.96, p = 0.50, q = 0.50, and e = 0.05, the result was n = 124.17, rounded to 125 participants. This formula justifies the required number, but it does not automatically make the actual procedure probabilistic. Since there is no complete selection list or evidence of random selection in the files, recruitment is transparently described as consecutive, based on availability, among eligible individuals encountered during fieldwork. Consequently, the estimates should not be generalized as if they were derived from a simple random sample. To select the aforementioned population, the inclusion criteria applied were: being 18 years of age or older, residing within the study area, and having prior knowledge of or exposure to information about the Cañariaco project, as well as voluntary acceptance and signing of the informed consent form.

No one was excluded due to illiteracy or disability. When necessary, the team conducted a guided reading, explained the alternatives without suggesting answers, and allowed the participant to indicate their choice. This adaptation reduces barriers, although any future replication should individually document the type of support used and apply formal accessibility standards.

Data collection was conducted exclusively in person using printed questionnaires. Researchers explained the purpose, the voluntary nature of participation, the confidentiality of data processing, and the right to withdraw. Written consent was obtained during the same meeting. Google Forms, Zoom, or email consent forms were not used. Responses were coded with identifiers E1 to E125, digitized in spreadsheets, and subsequently checked for missing or out-of-range values. The two databases provided contain 125 complete records and valid ratings (1 to 5) across the 30 articles.

According to the protocol described by the authors, the questionnaire and the explanation of the consent form were available in Spanish and Quechua (Inkawasi-Cañaris). Participants could request oral assistance in their preferred language. The analytical database does not record which version each participant used, nor does it include an indicator of their native language; therefore, it was not possible to compare responses by language, assess metric equivalence, or verify invariance between versions. Before a new request is made, the original texts of both versions, the cross-cultural translation and review process, and a field identifying the administration method should be included.

As shown in Table 1, the technique used was a survey, and two questionnaires developed by the team were employed, each with 15 items and a five-point Likert scale: 1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, and 5 = strongly agree. The instruments were reviewed by three experts: Carla Mena Nevado, Helen Alberca Guerrero, and Javier Salazar Ipanaqué. The previous version only reported overall reliability; in this revision, internal consistency was recalculated for each instrument and dimension.

For each five-item dimension, the equivalent descriptive ranges are: low = 5–11, medium = 12–18, and high = 19–25. Inferential analysis used continuous scores because categorization reduces information and statistical power. The literal text of the items is not included in the two provided databases; for scientific integrity, it must be incorporated from the original instrument as supplementary material, without being reconstructed or inferred from codes P1–P15.

The provided databases do not include age, sex, education, occupation, native language, or length of residence. Consequently, it is not possible to describe the sociodemographic composition of the sample or control for confounding factors using these covariates. This omission is explicitly stated and acknowledged as a material limitation. If separate data sheets exist, they must be integrated using an anonymous identifier before application; if they do not exist, it is inappropriate to fabricate percentages or attribute demographic representativeness to the study group.

In this scientific article, the survey was used as a technique and as an instrument 2 questionnaires were applied, which were validated by 3 experts in the field, these experts were Mg. Carla Mena Nevado, Mg. Helen Alberca Guerrero and Mg. Javier Salazar Ipanaque.

Internal consistency was estimated directly from the items using Cronbach’s alpha. For communication strategies, α = 0.937 was obtained, with adequate values for each dimension. For environmental perception, α = 0.838 was obtained; the environmental knowledge dimension showed α = 0.542, a value that warrants caution as it suggests item heterogeneity or low covariation. The attitudes and responsibility dimensions showed acceptable consistency. Reliability does not demonstrate construct validity; therefore, future versions should incorporate factor analysis and cross-cultural evaluation.

• • Data cleaning and verification of ranges, missing values, and consistency of sums. No missing data were identified in the analyzed variables.

• • Descriptive statistics using frequency, percentage, mean, standard deviation, median, range, and interquartile range.

• • Internal consistency using Cronbach’s alpha (overall and by dimension).

• • Bivariate associations using Spearman’s rho. Precision was estimated using percentile confidence intervals obtained from 10,000 bootstrap resamples.

• • Multiple linear regression with overall environmental perception as the outcome and clarity, channels, and feedback as simultaneous predictors. Due to non-normality of residuals and heteroscedasticity, the main inference used robust HC3 standard errors.

• • Collinearity was assessed using variance inflation factors (VIF), residual normality using Shapiro-Wilk, and heteroscedasticity using Breusch-Pagan and White.

• • Sensitivity analysis: minimum detectable correlation with 80% power and minimum detectable R² for three predictors. Post hoc power was not interpreted as proof of the null hypothesis.

Participation was voluntary, and written informed consent was obtained before administering the questionnaire. Objectives, procedures, foreseeable risks, confidentiality, and the right to withdraw without consequence were communicated. Records were anonymized using codes and used for academic purposes. The study was conducted in accordance with the César Vallejo University Research Ethics Code, version 02, approved by University Council Resolution RCPU No. 0659-2024-UCV, and underwent institutional review, as reported by the authors. The absence of direct identifiers reduces the risk of re-identification, although the complete omission of demographic variables limits the assessment of equity and representativeness.

One hundred and twenty-five questionnaires with 30 valid responses per participant were analyzed. No missing data were observed in the items, dimensions, or total scores. The available information allows for the description of the scales and the analysis of their relationships, but not the creation of a demographic table. Therefore, the results refer to “surveyed participants” and not to subgroups of age, sex, education, language, or occupation.

The overall Cronbach’s alpha for the environmental perception variable, recalculated from the items presented in Table 2, was 0.838, replacing the previously reported value of 0.856. This difference may be due to a different calculation method, rounding criteria, or the use of a different version of the database. For transparency and reproducibility, the manuscript reports the coefficient obtained directly from the data presented in Table 2 and from the deposited files.

According to the results presented in Tables 3 and 4, almost half of the participants rated the communication strategies as low, while only 8.8% rated them as high. Furthermore, the feedback dimension registered the lowest average among the dimensions evaluated. These results reflect an unfavorable assessment of the communication strategies as perceived by the respondents; however, they do not allow us to identify the causes of this assessment or to conclude that the company objectively implemented an ineffective communication strategy, since the variable evaluated the participants’ perceptions and not an objective measurement of the organization’s performance.

According to the results presented in Tables 5 and 6, the total environmental perception score was concentrated in the medium and high levels. This result indicates a favorable assessment of environmental perception within the applied instrument; however, it should not be automatically interpreted as evidence of sufficient technical knowledge, acceptance or rejection of the project, or a specific level of perceived risk, since these constructs represent conceptually distinct dimensions and were not directly measured by the instrument.

According to the results presented in Table 7, all confidence intervals included zero. In particular, the overall correlation estimate was very close to zero, and its confidence interval was compatible with both a small negative and a small positive association. Consequently, the results indicate that the sample did not provide statistically significant evidence of a monotonic association between the variables evaluated. Therefore, it is not appropriate to state that the null hypothesis has been “accepted” or “proven,” nor to conclude that there is a definitive absence of a relationship in the population, but only that, with the available evidence, no statistically significant association was detected.

According to the results presented in Table 8, a multiple regression model was estimated with overall environmental perception as the dependent variable and the three dimensions of communication strategies as simultaneous predictors. The variance inflation factors (VIF) were 2.68 for clarity, 3.40 for channels, and 2.86 for feedback, values that did not exceed the threshold of 5, although they showed a relatively high correlation among the predictor dimensions. Furthermore, the model assumptions showed that the residuals did not follow a normal distribution (Shapiro–Wilk, p = 0.004) and that heteroscedasticity existed, according to the Breusch–Pagan (p = 0.035) and White (p = 0.025) tests. Consequently, the model coefficients were reported using robust HC3 standard errors, which provides more reliable inferences in the face of these assumptions being violated.

The clarity coefficient indicates that, holding channels and feedback constant, one additional point in clarity was associated with 0.80 additional points in environmental perception. This result did not appear in the bivariate correlation and emerged when controlling for two highly related dimensions; therefore, it may reflect a conditional or suppression effect. The robust overall test was just above 0.05, and the explained variance was reduced. It should not be presented as causal evidence or as strong confirmation that increasing clarity will produce environmental changes.

According to the results presented in Table 9, the sample showed adequate sensitivity for detecting correlations of approximately 0.25 or higher, although it may not have had sufficient power to identify associations of smaller magnitude. Furthermore, the observed R² was lower than the approximate R² required to achieve a statistical power of 80%. Consequently, the overall lack of significance of the model can be attributed to the existence of a small effect size, the inherent variability of the sample, or the absence of an association between the variables. Therefore, with the available evidence, the results do not allow for a conclusive determination of which of these explanations is correct.

The results of the reliability analysis using Cronbach’s alpha coefficient showed appropriate values for both variables. The variable Strategic Communication obtained α = 0.935, and the variable Environmental Perception obtained α = 0.856. These values confirm the internal consistency and reliability of the instruments used.

According to the Kolmogorov–Smirnov test, none of the variables follow a normal distribution (p < 0.05), with values of 0.005 for Strategic Communication and 0.000 for Environmental Perception. For this reason, non-parametric tests were used in the correlational analysis.