2026-10-14T21:07:44Z https://www.scidb.cn/oai

10.57760/sciencedb.012um 2026-09-17T11:24:22Z

2026-09-17 Dataset of the Artificial Intelligence Literacy Assessment Questionnaire for Pre-service Teachers doi:10.57760/sciencedb.012um en This dataset was derived from the empirical investigation component of the project "ZhiFang AI Industry: Construction of an AI Literacy Framework for Pre-service Teachers and Empirical Research on Its Assessment," and consists of first-hand questionnaire data collected by the research team between July and August 2026. The survey targeted pre-service teachers (normal university students); the Pre-service Teachers' AI Literacy Assessment Questionnaire was distributed via the Wenjuanxing online survey platform, yielding 300 returned questionnaires. To ensure data quality, the research team applied three sequential cleaning rules to the returned data: removal of questionnaires with an excessively short completion time, removal of questionnaires in which the same option was selected consecutively, and removal of questionnaires containing obvious logical contradictions. After cleaning, 190 valid questionnaires were retained, giving a valid response rate of 63.3%. The dataset consists of these 190 valid samples. Data analysis was performed using IBM SPSS Statistics 31.0 and IBM SPSS Amos 26.0: descriptive statistics, exploratory factor analysis, reliability analysis and group difference tests were conducted in SPSS, while confirmatory factor analysis was conducted in Amos.The dataset is organised in tabular form, with each row representing one surveyed pre-service teacher, giving a total of 190 valid records with no missing values. The column labels fall into three categories. The first column is "ID", the respondent's number in the original questionnaire. The second column is "total score", the sum of the scores of all 18 items. The remaining 18 columns are item variables, named with three-digit codes (111, 112, 113, 121 … 163), each corresponding to an item in Part II of the questionnaire. Every three items belong to the same literacy dimension and correspond, in turn, to human-centred literacy, AI basic knowledge and cognition literacy, AI ethics and safety literacy, AI technology application and teaching integration literacy, AI teaching guidance and student development literacy, and human–AI collaboration and lifelong professional development literacy. All items use a five-point Likert scale, where 1 = "strongly disagree" and 5 = "strongly agree", and higher scores indicate a higher level of the corresponding literacy. As all variables are scale scores or coded variables, the dataset contains no physical measurement units; frequency information is counted in persons.The data cover the period from July to August 2026. The respondents were pre-service teachers, and the questionnaire was distributed online without geographic restriction; as social science survey data, the dataset involves no spatial resolution information. There are no missing values in the cleaned valid data. Three sources of error should be noted. First, the scale is self-reported, so the measurements reflect respondents' self-assessment and self-efficacy rather than their actual ability, and discrepancies are possible. Second, social desirability bias may lead respondents to give answers that appear more normatively appropriate on ethics- and responsibility-related items. Third, convenience sampling was used and 70.5% of the sample were female, which limits representativeness; conclusions should therefore be generalised with caution.The dataset contains two categories of files: survey data and statistical analysis outputs. The survey data files are: 清洗数据190.xlsx (Cleaned Data 190), in generic Excel format, containing all variable data for the 190 valid samples; and 完整222.sav (Complete 222) and 清洗数据222.xlsx (Cleaned Data 222), which are versioned files produced during the data cleaning process and allow that process to be traced. The statistical analysis output files are: 完整222.spv (Complete 222), an IBM SPSS Statistics output file containing the results of descriptive statistics, exploratory factor analysis, reliability analysis, independent samples t-tests and one-way ANOVA; and 新顺序.amw, 新顺序.amp, 新顺序.AmosOutput and 新顺序.bk1, which are Amos files used for the confirmatory factor analysis. Of these, the .amw file is the graphical modelling working file containing the path diagram specification of the four-factor model; the .amp file is the project file recording the analysis settings and parameters of that model; the .AmosOutput file is the analysis output containing model fit indices, standardised path coefficients and factor loadings; and the .bk1 file is an automatic backup of the working file. It should be noted that .sav, .spv, .amw, .amp, .AmosOutput and .bk1 are proprietary formats of the IBM SPSS family of software and require the corresponding software to open them (SPSS Statistics official download: https://www.ibm.com/products/spss-statistics; Amos is available through IBM official channels), whereas .xlsx files can be read with common spreadsheet software such as Microsoft Excel, WPS Spreadsheets or LibreOffice Calc. AI literacy; pre-service teachers; questionnaire ding wen qi PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank