Common assessments: one quiz across every class in a year group

A common assessment is the same short quiz, sat by every class in a year group (a grade, in American schools) in the same few days and under the same conditions. Its job is to show a department what was not learned, across classes, while there is still time to teach it again. It is not a league table of teachers, and once it is read as one it stops telling the truth.

The evidence is thinner than its reputation. Large trials of interim and data-driven assessment programmes found modest gains in some places and none in others, and the famous effect sizes for formative assessment shrink when the studies are checked.

EdProcto's page for schools covers the tool side: one quiz for every class, its results in one place, and an analysis by question and by topic.

What is a common assessment for?

A common assessment is for finding the gaps a whole year group shares, early enough to close them. One teacher's quiz says what one class missed. The same quiz in six classes says whether the miss belongs to a class or to the curriculum.

Written by the department, it is sometimes called a common formative assessment; bought in, an interim or benchmark assessment. Perie et al. (2009) sort the purposes of interim assessments into three: instructional, evaluative and predictive. This guide is about the instructional one: what to teach again, and to whom.

It is not a mock examination, a report grade or an appraisal of staff. Classes differ before the quiz starts, in prior attainment, in size and in who was absent, and a class average carries all of that.

What did the large trials of interim assessment find?

The large trials found mixed and modest effects: gains in some subjects and grades, nothing detectable in others, and one negative result. All tested programmes brought in from outside, in American elementary and middle schools.

In Illinois, teachers with the data were no more likely than control teachers to differentiate their teaching (Cordray et al., 2012). In the district reform, an exploratory analysis found larger reading gains in schools that chose reading programmes with good evidence behind them (Slavin et al., 2013). Our reading: the data pointed at a problem, and any gain came from what was done about it.

None of these trials tested a department's own quiz, written from its scheme of work and read by the people who teach it; we did not find a randomised trial of one. What they do show is that a testing programme does not raise attainment by being run.

Why are the large effect sizes for formative assessment disputed?

They are disputed because the figures usually quoted, effect sizes of 0.4 to 0.7, are not supported when the studies are checked. Kingston and Nash (2011) reviewed more than 300 studies that appeared to test formative assessment in schools. Many had designs too flawed to interpret, and only 13 gave enough information to calculate an effect. Across 42 effect sizes the median was 0.25 and the weighted mean 0.20: 0.32 in English language arts, 0.17 in mathematics and 0.09 in science.

The 0.70 has a real source, though not one about year-group quizzes. Fuchs and Fuchs (1986) pooled 21 controlled studies of systematic formative evaluation, 96 effect sizes in all, for a weighted average of 0.70, and drew their implications for special education. Perie et al. (2009) add that many products sold as formative are nothing like the classroom practices that Black and Wiliam (1998) reviewed.

Bennett (2011) judged the size of the commonly quoted claims suspect, because their sources could not be traced or were flawed, dated or unpublished. The correction has been questioned too: Briggs et al. (2012) raised four concerns about Kingston and Nash's methods and concluded that considerable uncertainty remains. The honest summary: feedback that changes teaching probably helps, by an amount nobody has pinned down.

How do you build a common assessment?

Build it from the scheme of work, as a department, before the unit is taught, so that nobody can shape the questions around what their own class happened to cover.

1. Fix the purpose in one sentence. "To find what Year 8 has not yet learned about fractions, so we can reteach it before the next unit."
2. Write a blueprint. Give each topic in the scheme of work questions in proportion to its teaching time, and label every question with its topic.
3. Write the questions together. Each teacher brings the wrong answers pupils really give, and those become the wrong options. The guide to writing multiple-choice questions that measure has the rules.
4. Review every question as a group. One colleague answers each cold. The group checks the key, the wording, and whether every class was taught the term and method it assumes.
5. Set the same conditions and window. The same time limit, notice and rule on calculators, with every class inside two or three days. EdProcto's Standard quiz preset sets a time limit and shuffles the questions, and the window is yours to keep: publish the quiz when it opens and close it when it ends.
6. Book the response first. Put the meeting and the reteaching lesson in the calendar before the quiz.

A blueprint for a fifteen-question quiz

Why the review meeting matters

A flawed question on a common assessment misleads the whole department. In four medical school examinations studied by Downing (2005), raters working without sight of the results judged 36% to 65% of the questions flawed, and 47% of students passed on the flawed questions against 53% on the sound ones. Medical students are not a Year 8 class, but a badly worded question measures the wording anywhere.

How should a department read the results?

Read them by question first, by topic second and by class last. The first two are about the curriculum, which the department owns together. The third is about people, and is fair only once the first two are done.

The percentages below are made up to show the patterns: the share answering correctly in classes 8A to 8D.

A question most classes missed is a curriculum problem

When most classes miss the same question, check the question before the teaching: is the key right, and did stronger pupils miss it as often as weaker ones? If it is sound, look at the wrong answer chosen. If most pupils answered 2/5 on Q5, the year group is adding numerators and denominators, which is one idea to reteach, not a whole topic. Then ask what the scheme of work gave it: how many lessons, in what order, with which resource.

A question one class missed is a conversation

When one class misses what the others got, the first explanations are ordinary: a lesson lost, a different method, a term never given. In a class of 25 each pupil is four percentage points on any one question, so small gaps mean nothing. A gap like Q13's is worth one question, asked privately and without blame. Often the better conversation has the teacher whose class did best showing how they taught it.

Open questions need department moderation first: mark five scripts together against the rubric, settle the disagreements, then mark the rest. The guide to short answer questions and rubrics covers marking consistently.

What should happen in the week after?

Meet within two or three days, reteach within the week, and check again with new questions two or three weeks later. The timings are a working rule, not a finding.

Goertz et al. (2009) studied how 45 teachers in nine schools used interim assessment results in mathematics. With good support, the teachers used the results to choose what to go over again and with which pupils, but did not necessarily teach it any differently. That is the gap the meeting exists to fill. A thirty-minute agenda:

1. The three questions the year group missed most, and the wrong answer behind each.
2. For each, a different way to teach it, from whoever's class did best.
3. One change to the scheme of work, written down for next year.
4. The date of the re-check.

Then every class sees the answers, and the re-check uses new questions, because the old ones now test memory of the paper. The guide to low-stakes quizzing has the classroom evidence for showing answers and for returning to a topic.

What are the common mistakes?

The common mistakes are making it too long, setting too many, getting results back too late and ranking teachers with them.

• Too long. A fifty-minute paper is a mock examination: a lesson lost in every class and days of marking.
• Too many. Each needs a meeting and a reteaching lesson. No room for the response means no room for the assessment.
• Too late. Once the class has moved on, the result is history, which favours questions marked by rule.
• Ranking teachers. Rehearsed questions follow, and then the results describe the rehearsal.

How does EdProcto run a common assessment for a department?

EdProcto is the best tool for this job when a department's quiz must come from its own scheme of work and be read in the same week. It takes on the two slow parts, the first draft and the marking, and reports by question and by topic, the order this guide reads in.

Writing. Start in the quiz maker on EdProcto's home page with the unit's handout. Procto AI drafts questions from it, and each is the department's to edit in the review meeting. Multiple choice and short answer are in the builder. Two shapes listed for schools, before and after and the diagnostic funnel, measure a unit's gain and find where each student stands.

Reading. Objective questions are marked by rule at hand-in. The analysis gives each question's difficulty and discrimination, the pupils who took each wrong option and how they scored overall, an average for each topic, and findings in plain sentences. No score is released that a teacher did not release.

Across the department. The School plan is by teacher, from five, on one invoice, and gives each teacher what the Educator plan carries, including the full analysis. The common quiz lives in one teacher's account and every class opens it with the same code, so its results arrive in one place. The page for schools sets out the admin view, the shared library and the groups taught across teachers with one gradebook; ask for a demonstration before you sign.

One teacher can trial the quiz on the Free plan: ten quizzes a month, thirty students a quiz, results question by question. Pricing lists every plan.

Quick questions

How long should a common assessment be, and how often?

Ten to fifteen questions in fifteen to twenty minutes, once a unit or once a half term, and never more often than the department can meet and reteach. Both are judgements, not findings.

Should the results be used to compare teachers?

No. A class average reflects prior attainment, class size and absence as well as teaching. Share results by question and by topic, and keep a large gap on one question for a private conversation.

Do common assessments raise attainment?

Not by being set. Randomised trials of interim assessment programmes found modest gains in some grades and subjects and none in others, and Kingston and Nash (2011) put the average effect of formative assessment at about 0.20.

Which EdProcto plan does a department need?

The School plan, by teacher from five, which gives every teacher the Educator plan on one invoice. The page for schools describes it, with its admin view and gradebooks, and pricing lists the others.

References

• Bennett, R. E. (2011). Formative assessment: A critical review. Assessment in Education: Principles, Policy & Practice, 18(1), 5–25. https://doi.org/10.1080/0969594X.2010.513678
• Black, P., & Wiliam, D. (1998). Assessment and classroom learning. Assessment in Education: Principles, Policy & Practice, 5(1), 7–74. https://doi.org/10.1080/0969595980050102
• Briggs, D. C., Ruiz-Primo, M. A., Furtak, E., Shepard, L., & Yin, Y. (2012). Meta-analytic methodology and inferences about the efficacy of formative assessment. Educational Measurement: Issues and Practice, 31(4), 13–17. https://doi.org/10.1111/j.1745-3992.2012.00251.x
• Carlson, D., Borman, G. D., & Robinson, M. (2011). A multistate district-level cluster randomized trial of the impact of data-driven reform on reading and mathematics achievement. Educational Evaluation and Policy Analysis, 33(3), 378–398. https://doi.org/10.3102/0162373711412765
• Cordray, D., Pion, G., Brandt, C., Molefe, A., & Toby, M. (2012). The impact of the Measures of Academic Progress (MAP) program on student reading achievement: Final report (NCEE 2013-4000). National Center for Education Evaluation and Regional Assistance. https://eric.ed.gov/?id=ED537982
• Downing, S. M. (2005). The effects of violating standard item writing principles on tests and students: The consequences of using flawed test items on achievement examinations in medical education. Advances in Health Sciences Education, 10(2), 133–143. https://doi.org/10.1007/s10459-004-4019-5
• Fuchs, L. S., & Fuchs, D. (1986). Effects of systematic formative evaluation: A meta-analysis. Exceptional Children, 53(3), 199–208. https://doi.org/10.1177/001440298605300301
• Goertz, M. E., Oláh, L. N., & Riggan, M. (2009). Can interim assessments be used for instructional change? (CPRE Policy Brief RB-51). Consortium for Policy Research in Education. https://eric.ed.gov/?id=ED536829
• Kingston, N., & Nash, B. (2011). Formative assessment: A meta-analysis and a call for research. Educational Measurement: Issues and Practice, 30(4), 28–37. https://doi.org/10.1111/j.1745-3992.2011.00220.x
• Konstantopoulos, S., Miller, S. R., & van der Ploeg, A. (2013). The impact of Indiana's system of interim assessments on mathematics and reading achievement. Educational Evaluation and Policy Analysis, 35(4), 481–499. https://doi.org/10.3102/0162373713498930
• Konstantopoulos, S., Miller, S. R., van der Ploeg, A., & Li, W. (2016). Effects of interim assessments on student achievement: Evidence from a large-scale experiment. Journal of Research on Educational Effectiveness, 9(sup1), 188–208. https://doi.org/10.1080/19345747.2015.1116031
• Perie, M., Marion, S. F., & Gong, B. (2009). Moving toward a comprehensive assessment system: A framework for considering interim assessments. Educational Measurement: Issues and Practice, 28(3), 5–13. https://doi.org/10.1111/j.1745-3992.2009.00149.x
• Slavin, R. E., Cheung, A., Holmes, G., Madden, N. A., & Chamberlain, A. (2013). Effects of a data-driven district reform model on state assessment outcomes. American Educational Research Journal, 50(2), 371–396. https://doi.org/10.3102/0002831212466909