Tips for Importing Exam Files in ExamEval

Supported File Types and Formatting
ExamEval uses AI to process exam files, so questions and answers can be in any format. There are no requirements for how questions are numbered, answer choices are listed, or spacing between questions.
ExamEval supports the following file types:
- Word (DOC, DOCX, ODF)
- Rich text (RTF)
- Plain text (TXT)
Processing Limitations
ExamEval converts exam files into "plain" text, meaning text without formatting (bold, italics, coloring) and without images. Information presented in tables may not be correctly converted to plain text.
If a question's context depends on formatting or an image, ExamEval will not capture the full context during analysis. This can lead to falsely identifying a question as "flawed" or not flagging a question when a flaw truly exists.
Tips for Optimal Results
Include Correct Answer Indicators
Exam analysis is significantly improved when the correct answer is provided. Including an indicator of the correct answer allows for greater accuracy in identifying miskeyed answer choices or 'single best answer' questions that actually have more than one correct answer.
A correct answer can be indicated using a symbol (such as an asterisk or ✓) or a brief explanation following the question. Bolding, italics, or other formatting to indicate a correct answer is lost when converting an exam file to plain text and cannot be "seen" by the AI during analysis.
Clean Up Your Files (Optional)
While ExamEval's AI is designed to extract questions and answers and ignore extraneous information from exam files, in some cases, extraneous text may be incorrectly identified as part of an exam question. If this occurs, consider removing the following content from your exam file before uploading:
- Headers and footers - Page numbers, dates, and institutional headers
- Instructions and disclaimers - General exam instructions, academic integrity statements
- Non-question content - Course syllabi, reference materials, or unrelated text
Understanding Analysis Limitations
Question Formats and Analysis Quality
ExamEval is primarily designed to analyze item-writing flaws for multiple choice questions, particularly single best answer and select all that apply formats. Fill-in-the-blank, short answer, and essay questions receive limited analysis that focuses mainly on grammar and spelling clarity.
Questions with other formats, such as matching, ordering, "hot spot" clickable images, or drag-and-drop, may not be analyzed correctly and could trigger false positive flaws.
When Visual Elements Are Present
If your exam includes questions with images, charts, tables, or diagrams, ExamEval will likely falsely "flag" these questions as being flawed because images are not included during AI processing and table formatting is lost.
In these cases, faculty are encouraged to ignore these false positives and conduct a more thorough manual review of the question rather than relying on AI analysis.