Mediating effect of use perceptions on technology readiness and adoption of artificial intelligence in accounting Data collection
Question:
Mediating effect of use perceptions on technology readiness and adoption of artificial intelligence in accounting
Data collection and research sample
The study was conducted in two universities in the United States and was related to accounting students’ perceptions of TR and AI technology acceptance (or technology adoption). The criteria for the institutions selected were (a) is reputable and accredited by the Western Association of Schools and Colleges, (b) offers degree-bound under- graduate accounting programs, and (c) supplies the accounting and auditing industry in the United States with accounting graduates. The institutions selected for the study were not randomly selected, which is one of the limitations of the study. The institutions were selected based on relevance to the above-mentioned criteria and geographical con- venience for the researcher to perform the study. The inclusion of multiple sites increases the sample size in an attempt to generalize across the population. As the sample sizeincreases, the standard error decreases (Creswell, 2014). Finally, studying multiple insti- tutions added increased anonymity to the participants, which may be desirable for the institutions when examining the topic of AI TA by accounting students. However, even if only one site was selected, the standard error may still be sufficiently small as long as the researcher achieves the required sample size for a correlation study. Accord- ingly, the study could still be generalizable since most universities offer similar topics as part of their standard accounting curriculum.The target population for this study comprised all 824 accounting students (N = 824) from two universities in the United States. A computer-administered survey method was used to collect primary data from respondents using Qualtrics. The survey was carried out with accounting students after being granted permission by the Institutional Review Boards (IRB) at both institutions. The survey instrument, provided in the Appen- dix, was uploaded onto the Qualtrics platform, and a link to the questionnaire was gen- erated, captured, and published on 16 August 2019. The researcher decided to open the survey link when students were two weeks into the fall semester and were more likely to participate in the study. The link was then distributed to the target population at each of the two research sites via e-mails from the chair of the accounting department and the dean of the college of business and public management, respectively.The researcher decided to close the survey link when 100 responses had been gathered; a sufficient sample size (> 55) to yield a power of 0.8 (80%) based on G*Power (Faul et al., 2009). On 6 September 2019, the researcher closed the survey after a total of 105 responses were obtained, three of which were blank and removed from the survey. In the initial phase of data analysis, the researcher analyzed data using Mahalanobis’s, Cook’s, and Leverage’s tests to identify outliers. One observation was identified as an outlier in both Cook’s and Leverage’s tests and was, therefore, removed from the dataset. The outlier observation exhibited medium to high values in TR, PU, and PEOU but a very low score on TA, which logically contradicted a direct and positive relationship among variables. Accordingly, 101 responses (n = 101) were finally retained for the purpose of data analysis.
Sample size
To determine whether 101 participants were an adequate sample size for a linear regression analysis, a G*power analysis was conducted. It was determined that a sample size of 55 would be sufficient to yield a power of 0.8 (80%). G*Power (Faul et al., 2009) is a stand- alone power analysis program for many statistical tests, such as correlation and regressions, commonly used in the social, behavioral, and biomedical sciences. Accord- ingly, a sample size of 101 (> 55) was deemed sufficient for regression and mediation ana- lyses. Note that mediation analysis is a form of multivariate linear regression.
Q.1.4 Identify the data collection methods used in each article.
Q.1.4.1 | Using examples from each article, identify the data collection methods used in each study. Motivate your answer with examples from each article as well as theory from various sources. | |
Q.1.4.2 | For each article, using appropriate theory, argue in favour of an alternate data collection method that aligns with the same paradigm/tradition and research question. |