The Relationship of Formative and Summative Assessment

In an October 13, 2006 blog entry prompted by an American Federation of Teachers policy forum Kent Williamson asks what assessment measures teachers use that provide them information to improve instruction. At the Spelling Commission forum on November 22, 2006 panelists Peg Miller and Peter Ewell made a distinction between the kind of formative assessment that provides such helpful information and the kind of summative assessment, sometimes labeled evaluation, that rates performance at a certain point in time. I am encouraged that we are beginning to distinguish between formative and summative assessment, not to affirm one over the other but to accent that each serves a particular purpose, formative primarily to improve teaching and learning and summative primarily to answer accountability needs.

At the same time that we make that distinction, however, we might also think how doing each kind of assessment might affect the other kind. To return to the AFT forum Kent referred to, listen to points made by Paul Barton in a forum presentation labeled "'Failing' or 'Succeeding' Schools: How Can We Tell?" Barton makes four points about accountability data, the kind of data most often generated by summative assessments.

First, Barton contends that current practices ignore basic standards of accountability because curricula and tests are not yet aligned. Test scores used for accountability are invalid if alignment is not in order. My take from this point: Pedagogy, curriculum, and formative assessment need to track with summative assessment.

Secondly, Barton states that a series of snapshots of students in different years does not measure what is learned by a student in a certain school year. Barton recommends administering the same test at the beginning and the end of the school year. My take on this point: Formative assessments can be used to track progress during that school year so that students can be helped to make more progress between summative assessments .

Thirdly, Barton says gains measured during the school year should be transparent to everyone, especially teachers and parents. Transparency would be supported by having student identifiers to track students from grade to grade or by stretch tests that cover several grades worth of work but are taken each year. Tests, however, are really not needed every year. Samples, rotating testing, or testing on an unannouced basis would free time for more diagnostic testing (formative assessment), which research shows improves instruction. My take on this point: A system of testing periodically can serve accountability while honoring more frequent formative assessment that serves teaching and learning.

Fourthly, standards need to be set for how much gain is expected in a year. Teachers need to say what is typical at a low end and at a high end. We can still have high expectations and disaggregate by subgroup under this standards system. My take on this point: Hurray that teachers are identified as the professionals who should set standards.

Each of Barton's points, it seems to me, recognizes the (potential) interaction of formative and summative assessment.

Do you have different takeaways from Barton's points? Do you have examples of how formative and summative assessments are functionally mutually well in your school? Do you have suggestions about how they might do so?

Barbara