Exploring ChatGPT’s capabilities in solving accounting standards problems: the case of IAS 37
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Using a quasi-experimental method and content analysis as a technique, this study tests ChatGPT, in its version 4, by assessing its textual characteristics and overall understanding regarding the recognition criteria of provisions under International Accounting Standards (IAS) 37, as issued by the International Accounting Standards Board (IASB). For this purpose, it uses a set of questions (input) from the IASB's illustrative examples to compare the answers (output) from IASB and ChatGPT in two distinct strategies: with and without prompting. The findings indicate that ChatGPT’s answers are wordier, have higher magnitude levels, and are more predominantly inserted in Business and Finance. The no-prompting strategy is globally more negative and subjective, while the prompting one improves the answers’ focus and readability, also presenting more diverse tones in its textual characteristics, similar to what was found in the IASB's answers. However, some answers were not globally accurate in both strategies. These findings provide insights into how ChatGPT, as one of the most disseminated artificial intelligence tools, can be used by accounting professionals and educators, being aware of the potential risks and benefits from both strategies underlying this experiment. Then, by considering those aspects, practitioners, including accountants and managers, but also investors can use it to understand the matters and issues under assessment in a given situation, as well as the sources to consider when preparing financial statements or making a decision. Academics can also use it to open up discussions and promote students’ critical thinking skills in a classroom environment.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it