Research on Exam Paper Format and Structure - Study on Font, Response Space, Stimuli, and Question Design

Oct 20, 2026

Research on font choice, response space, stimuli, and layout shows that exam format materially affects candidate performance. Evidence indicates some typesets support more accessible, accurate, and more fully developed answers, while combined booklets generally improve marks yet may restrict space for long-answer responses. Stimulus design and information placement influence how candidates interpret questions and recall details. Question type, order, and arrangement (including easy-to-hard sequencing) shape assessment effectiveness and anxiety effects, and gender-related patterns may emerge.

LanguageEnglish

CategoryExam

FormatPDF

Pages15

Reading Time38 min

AccessFree to download

Background and evidence

Typography and readability

Response format and space

Stimuli and information placement

Question type, order, and performance factors

Study method and participants

Design and recruitment

Materials and stimulus materials

The research literature regarding font, response space, stimuli, and layout demonstrates that exam format can affect a candidate’s performance. Research has shown that some typesets are better than others. Arial and Comic typeface are preferred to Times New Roman for onscreen reading and it has been argued that these fonts are more accessible to children as they resemble children’s handwriting and are less formal than Times New Roman (Bernard, Chaparro, Mills, & Halcomb, 2002; 2003) . The format of the provision for candidates’ responses has also been shown to affect exam performance. In general, candidates achieve better marks, write longer and fuller answers, and are more precise when using combined booklets compared to separate booklets. Candidates have reported that combined booklets encouraged them to answer questions and provided some guidance regarding the length and depth of the expected response (Crisp, 2008) . However, candidates have reported running out of space for longanswer questions when using combined booklets (Crisp 2008) . The logical solution to this problem is to simply provide more lines; yet providing more lines may result in some candidates wasting time writing longer answers than required (Crisp, Sweiry, Ahmed, & Pollitt, 2002) . Stimuli can influence candidates’ interpretation of the questions (Fisher-Hoch, Hughes, & Bramley, 1997) . Images have many benefits, such as making the abstract more concrete, motivating students to respond, and making the paper seem less daunting (Crisp & Sweiry, 2006; Johnson, 2004) . On the other hand, Pollitt and Ahmed (2000) found that candidates lost marks by trying to make images relevant for questions where the image was intended to provide context but was not required in the candidate’s response. The location of information on the page is significant; research into memory suggests that information which is accessed first and last is most reliably recalled from short term memory (Feigenbaum & Simon, 1962; Murdock, 1962; Wiswede, Russeler, & Munte, 2007) . With this in mind, important information may be best placed at the beginning, with the task instruction placed at the end. Crisp’s (2008) interviewees reported that the proximity of the response space to the questions helped them respond specifically to the question asked. Having the question close to where candidates write their answer reduces the need to remember the question, which may explain why Crisp’s participants found this to be helpful. A paper’s structure is a function of the type and order of the questions within it. Certain types of question are more likely to be successful at assessing some skills than others. For example, essay-based questions are historically thought to be better at assessing divergent skills, whereas multiple-choice questions are considered better at assessing breadth of subject knowledge (O’Donovan, 2005) . If a paper is constructed entirely from one type of question, there is a danger that the paper will have a systematic bias. Research has demonstrated sex differences in responses to some question types. Males have been observed to do better than females in multiple-choice objective exams, while females do better than males in open-ended exams (Murphy, 1980) . In some cases, gender bias may be associated legitimately with the subject content rather than question type (e.g. Maths) (Penner & Paret, 2008) . Equally, multiplechoice questions have evolved to test a variety of skills and now come in many forms (Stringer, 2009a) . The structure of the paper will vary with question order and there is a consistent finding in the literature that an easy-to-hard question arrangement results in better performance than a hard-to-easy question arrangement (Leary & Dorans, 1985) . Anxiety may play a role in this, in that difficult questions at the beginning of an exam could heighten anxiety levels and reduce exam performance (Mulkey & O’Neil, 1999) . Equally, candidates answer questions sequ

FAQs

How does font and onscreen readability influence exam performance?

Research cited in the text indicates that some fonts are preferred for onscreen reading and can be argued to be more accessible to children than Times New Roman, influencing how candidates read and engage with items.

Why can combined booklets improve marks but also create a trade-off?

The text reports that combined booklets lead to better marks, longer and fuller answers, and greater precision, and that they encourage answering. However, candidates also report running out of space for long-answer questions in combined formats.

What role do question order and layout play in candidate outcomes?

The literature summarized notes that an easy-to-hard question arrangement tends to produce better performance than hard-to-easy. It also highlights that anxiety can rise when difficult questions appear early, while proximity of response space to questions can reduce memory demands and help candidates answer specifically.

Oct 20, 2026

Uploaded by Riley West