Struggling and juggling: a comparative analysis of assessment loads across different universities

Tansy Jessop

    Research output: Contribution to journalArticlepeer-review


    In spite of the rising tide of metrics in UK higher education, there has been
    scant attention paid to assessment loads, when evidence demonstrates that
    heavy demands lead to surface learning. Our study seeks to redress the
    situation by defining assessment loads and comparing them across research
    and teaching intensive universities. We clarify the concept of ‘assessment
    load’ in response to findings about high volumes of summative assessment
    on modular degrees. We define assessment load across whole undergraduate
    degrees, according to four measures: the volume of summative assessment;
    volume of formative assessment; proportion of examinations to coursework;
    number of different varieties of assessment. All four factors contribute to
    the weight of an assessment load, and influence students’ approaches
    to learning. Our research compares programme assessment data from
    73 programmes in 14 UK universities, across two institutional categories.
    Research-intensives have higher summative assessment loads and a greater
    proportion of examinations; teaching-intensives have higher varieties of
    assessment. Formative assessment does not differ significantly across both
    university groups. These findings pose particular challenges for students in
    different parts of the sector. Our study questions the wisdom that ‘more’ is
    always better, proposing that lighter assessment loads may make room for
    ‘slow’ and deep learning.
    Original languageEnglish
    Pages (from-to)1-10
    JournalAssessment and Evaluation in Higher Education
    Issue number1
    Publication statusE-pub ahead of print - 18 Apr 2018

    Fingerprint Dive into the research topics of 'Struggling and juggling: a comparative analysis of assessment loads across different universities'. Together they form a unique fingerprint.

    Cite this