Accelerated CREP project number: AC1938. *This component contains all datasets (anonymized raw dataset, edited dataset, final dataset and dataset in long-format for our analyses), codebooks, R-scripts, main results, (full) analysis report and a description of all files.* **DATA** File *JTB_anonymized raw dataset* (N = 218) presents translated version of the dataset, all cases are included. File *JTB_anonymized dataset_exclusion1* (N = 154) is a set of all valid responses - responses with no completed vignette were excluded. To create file *JTB_final dataset* (N = 121), we further excluded participants based on reasons listed in the preprint. One participant was underage, one participant stated he/she participated in a similar study and one participant proved insufficient language proficiency. Any participant correctly and explicitly articulated knowledge of the specific hypotheses or specific conditions of the study. Finally, thirty participants failed to answer one or two comprehension questions correctly and were excluded: A) six participants in Darrel vignettes (1 in knowledge control cond., 4 in Gettier case cond., 1 in ignorance control cond.), B) seventeen participants in Gerald vignettes (4 in ignorance control cond.*, 11 in Gettier case cond., 2 in ignorance control cond.), C) eleven participants in Ema vignettes (1 in knowledge control cond., 3 in Gettier case cond., 7 in ignorance control cond.). Total of 19,87 % of cases were excluded based on this criteria. **HYPOTHESES AND PLANNED ANALYSIS** We used mixed linear model analysis with gender, age, education and experimental condition (JTB case, Gettier case, or Ignorance case) as fixed effects, and participant's ID and vignette type (Gerald, Darrel, or Emma) as random effects. We tested the hypothesis that in the Gettier case condition participant will score higher in knowledge attribution than in ignorance case condition and not higher or lower than in JTB case condition. We also tested the hypothesis that the reasonableness attribution changes depending on the type of the case. We considered we find support for our hypotheses based on significance (p < .05) of specific coefficients in the mixed linear model. **MAIN RESULTS** In our final model we included a random-intercept for participants and fixed effect for following predictors: experimental condition, type of vignette and demographic variables (i.e., gender, age and years of education). This model explained 22,8% of the variability in knowledge attribution. With respect to our hypotheses, participants in the Gettier condition scored significantly higher in knowledge attribution than participants in Ignorance control condition (b = -23.55, t(234.75) = -5.11, p< .001), supporting our original knowledge attribution hypothesis. However, contrary to our assumptions, Gettier case condition did not score equally or higher to knowledge control condition, but significantly lower (b = 15.97, t(250.63) = 2.89, p= .004). Based on the same multilevel analysis model, but with reasonableness as the dependent variable, our results do not support the hypothesis that reasonableness attribution changes depending on the type of the case. Reasonableness attributed to knowledge control condition was not higher than in Gettier case condition (b = -0.37, t(247.07) = -0.16, p = .877). In case of ignorance control condition, reasonableness score was lower compared to Gettier case condition, but this difference was also insignificant (b = -3.00, t(233.22) = -1.49, p = .138). In conclusion, we did find support that Gettier condition makes people attribute more knowledge to the person than in ignorance condition, but less than in knowledge condition. Regarding the reasonableness attribution, we found no significant differences between conditions. Please note that our conclusions are greatly limited by unmet assumptions for multilevel analysis (For detailed analysis and limitations, please *See* a file *analysis report*.) **DESCRIPTION OF FILES** *"JBT_final-dataset"* : A dataset (N = 121) after exclusion of observations based on the exclusion criteria stated in the Preprint. *"Codebook_JBT_final-dataset"* : A codebook for the dataset "JBT_final-dataset". *"code_data_transformation"* : A R-script that takes "JBT_final-dataset" as an input and transforms this wide-format dataset (i.e., 3 observations for each participant in one row) to a long-format dataset "data_long" (i.e., three rows for each participants - separate row for each observation within participant). The output of this R-code is the dataset "data_long" with selected variables necessary for our further analyses. *"data_long"* : A dataset in long format with a selection of variables for our further analyses. *"Codebook_data_long"* : A codebook for the dataset "data_long". *"code_analysis"* : A R-script for our analyses which takes the dataset "data_long" as input. It is divided into 3 main parts - descriptive statistics, multilevel models and tests of assumptions. *"Analysis-report"* : A comprehensive report of our analysis, which presents descriptive statistics, main findings based on the multilevel model and a discussion of the present limitations.