I have just finished reading Evaluation to Improve Learning by Benjamin Bloom, George Madaus, Thomas Hastings. It was written in 1981 and contains some fascinating connections to modern-day proficiency learning in the United States and especially in Vermont.
I have written before about the connections between modern proficiency learning (as we call it in Vermont) or competency learning (as they call it elsewhere) and earlier incarnations, including William Spady’s outcome-based education and Bloom’s mastery learning, the original version of the whole approach. As I’ve proposed in the past, both mastery learning and outcome-based education each swept across the ed landscape like wildfires, only to burn out in nearly as short a time. In both cases the causes were a mixture of philosophical and political, and what’s particularly interesting is the way that it does seem to me as though modern proficiency learning is somewhat here to stay, at least in my home state. I often have cause to wonder why this is, and reading this book by Bloom and company reminds me of a very simple fact: there’s something really, really appealing about this approach.
And that’s what first struck me in reading this book: it’s hard not to read it and fail to be won over. Bloom, et al make a powerful case for mastery learning, and it all starts with a basic, powerful premise. Immediately Bloom (and though he is just a co-author, for simplicity’s sake I’ll merely refer to his name) draws a clear contrast between his philosophical approach and those of traditional American education. He terms the first section, “Selection versus Development.” He writes of the past, “Much of the energy of teachers and administrators was devoted to determining the students to be dropped at each major stage of the education program” (Bloom, 1981, p. 2). This Platonic view of education as a sifting and sorting mechanism is something that Bloom regards both as immoral and, implicitly siding with Dewey, inefficient. Bloom contrasts this with his own vision of education, which is one that prizes not selection but individual development. “The underlying assumptions” of this view, he writes, “are that talent can be developed by educational means and that the major resources of the schools should be devoted to increasing the effectiveness of individuals rather than to predicting and selecting talent” (Bloom, 1981, p. 3).
This reminds me somewhat of Dewey, although much more technocratic/scientific and behaviorist than Dewey’s more naturalistic/growth-minded approach. Additionally, Bloom seems to be restricting himself largely to academic learning; although perhaps he agrees with it, there is no mention of Dewey’s capacious sense of learning being possible outside the classroom. And Dewey, like his modern day Pragmatist heirs, has little interest in formal evaluation – something that Bloom, coming from much more of a scientific-measurement background, devotes considerable time to. Still, Bloom’s view of the maximum development of individual students hews much closer to Dewey’s developmentalism than to Plato’s hierarchical Kallipolis.
And as in some of Dewey’s writing (particularly in Human Nature and Conduct), there’s a surprisingly moral bent to Bloom’s arguments in favor of mastery learning, too. He writes that the expectations in the traditional system on the part of teachers and curriculum designers is that many students simply won’t grasp the material; this, he writes, is not only wasteful of human talent, but serves to “alienate youth from both school and the community at large” (Bloom, 1981, p. 51) to such a great extent that “no society can tolerate it for long” (Bloom, 1981, p. 51). Later he talks of how the “system has been rigged” (66) in teaching scores of students that “they are inadequate, rather than the system of grading or the instruction” (66) and even arguing that the continued success students experience under mastery learning “can be one of the more powerful sources of mental health” for students (Bloom, 1981, p. 66). This is certainly a powerful appeal.
And this is really the crux of the importance of mastery learning, for Bloom – the notion that “given sufficient time and appropriate types of help, 95 percent of students (the top 5 percent plus the next 90 percent) can learn a subject with a high degree of mastery” (Bloom, 1981, p. 54). Put in different terms, he writes, “we are convinced that the grade of A as an index of mastery of a subject can, under appropriate conditions, be achieved by up to 95 percent of students in a class” (Bloom, 1981, p. 54). He makes a dramatic conclusion, surely one of the more optimistic and hopeful visions one can imagine an educational researcher making. Once again, this seems very different from a Plato – who wished to identify what a citizen’s talent was and then attempt to maximize that talent, or even of a Dewey, who merely wished to balance the social and political standing of vocational work to make it equivalent in prestige and educational focus to intellectual work. Yet there was some of Plato in Dewey – he did, it seems to me, wish to see each person identify their talent and run with it; if anything, he surely wanted less traditional subject matter, traditional testing, and less attempt to formalize learning experiences. Bloom, meanwhile, has no quarrel with traditional testing and traditional academic material – he believes it is masterable by (and, by inference, desirable for) nearly any student.
Nor does Bloom believe this high level of achievement is accomplishable by remade attitudes on the part of educators, nor by deep levels of engagement deriving from Deweyian/Dennis Litky-ian “big picture” thinking (learning outside the classroom in non-traditional ways). Bloom says little, in fact, about engagement at all. Instead, Bloom’s focus is on showing educators (and students) what is possible by leveraging clarity of objectives, formative feedback to aid precise, individualized instruction, and what he calls the “corrective process” to provide struggling students more time to learn the material.
And it’s the *rate* of learning that’s important. Bloom’s belief in each student’s ability to learn is given grounding and succinct summarization in JA Carroll’s formulation, quoted in the book, that aptitude is not a the complexity of material one can master, but “the amount of time required by the learner to attain mastery” (Bloom, 1981, p. 53). In other words, if given the same type of instruction in schools, students of differing abilities will attain vastly different results; if given different types of instruction and, more importantly, different amounts of time, “all students can conceivably attain mastery of a learning task” (54).
Bloom’s ideal, as he wrote about back in the 1960s, first describing mastery learning, was that of an individual tutor: give each child a tutor who can individually work with them and that child is capable of achieving at just as high a level as is required. That said, tutoring is expensive and time-consuming, so Bloom searched for ways to approximate this individual knowledge of student needs in a more traditional academic setting, rightly intuiting that although in many cases it was philosophical objections (the Platonic notion of education as sorting, or even the Deweyian notion of finding one’s niche or passion) that got in the way, too often it was the harsh realities of too many students and not enough time that made teachers fold up shop and move on to the next lesson even as many students still struggled to understand the core concepts for highly individualized reasons.
To be honest, Bloom’s tactics for working through the “corrective process” that he wrote about back in the late 1960s always struck me as a little bit unrealistic, and he reiterates some of them in his book. Small study groups of students teaching each other seems to be his favorite; this is the sort of activity all too likely, in a real-live class to go sideways. At the very least, given Bloom’s focus on giving and regiving similar, parallel assessments to students, the notion of chasing an 80 or an 85% on easily-corrected multiple choice quizzes – which is sort of what Bloom sometimes sounds like he’s advocating – can certainly help one understand the ways in which mastery learning was taken in the wrong direction. Talk to anyone who was in school in the early 1980s under a “mastery learning” program, and they’ll remember a curriculum composed of worksheet after worksheet, all bending toward “mastery” of rote memorization, easily gradable multiple choice quizzes, all of it getting further and further away from the depth of learning that the father of the Taxonomy so prized. The seeds of the movement’s downfall were surely there from the start – and you can see them on display in this book.
Still, the component of Bloom’s program that seems to me the most promising – and the one that I believe has shaped modern proficiency-based learning the most – is the idea of using formative assessment to provide feedback to both teacher and learner about a student’s needs and progress. Bloom compares traditional quizzes, the threat of which is often designed to motivate the learner, and the recording of which grade, counting on the final grade, is also meant to motivate, with the opportunity for stakes-free practice – and for the diagnosing of individual mistakes in learning on the part of the instructor. The metaphor he uses is apt: a traditional quiz is a thermometer (which merely measures the temperature – or degree of learning); the formative assessment is a thermostat (which both measures and provides a corrective to the existing temperature). This is closely aligned to another apt metaphor I heard about formative assessment: when the chef tastes the soup, it’s formative; when the guests taste the soup, it’s summative.
No matter how implausible, rote, or fanciful Bloom’s specific prescriptions for formative assessment are, the idea of using formative assessment to diagnose student mistakes and – this is important – to inform instruction (both on the level of further group instruction, and on the level of individual instruction to specific students) is certainly an important idea. I think the first time I read about this was in John Hattie’s famous meta-study, Visible Learning – in which Hattie makes a powerful, powerful argument in favor of, when you think about it, basically the same concepts that Benjamin Bloom was telling us all to do back in 1967: make learning visible with students by clearly defining what they are expected to do, then employing formative assessment to derive feedback – which Hattie always stressed was often comprised of feedback from students to teacher (about where they are in their learning, and, in a sense, how we teachers are doing with reaching their students). Bloom believes that few teachers use quizzes to inform instruction. “The primary change” that he suggests for the use of formative assessment is “that it be directed to yielding information which can be used to alter instruction or to review those ideas on which students are having great difficulty” (Bloom, 1981, p. 173). Often Bloom sounds just like Hattie: “Learning is a process which can be observed and evaluated as it is taking place” (Bloom, 1981, p. 176).
Now, will this system of – perhaps ungraded – formative feedback provide motivation for students, the same way that, he admits, old-school quizzes did? To his credit, Bloom addresses this question. “It is to be noted,” he writes, “that some students may reduce their learning effort when their test results are not graded” (Bloom, 1981, p. 171). To remedy this, he offers one suggestion: give students the option of their grade being determined either by just the summatives, or by the summatives combined with the formatives, whichever score is higher. It’s an interesting, if implausible, solution, but more broadly it should be said that Bloom again, in his section on grading, makes the case that the increased chance of mastering material will be motivating enough for students under his system: “The repeated evidence of mastery is a powerful reinforcement which will help ensure that the student will continue to invest the appropriate effort and interest in the subject” (Bloom, 1981, p. 170). In fact, he questions whether grading really does motivate students: “It is less certain that grading students on quizzes or formative evaluation is a useful reinforcement” – because students who feel themselves stuck in a C-level grade will be less likely to study hard enough to break out of this cycle. An interesting point – and not unconvincing.
“For this reason,” Bloom writes, “we believe formative evaluation should simply inform the students whether they have or have not mastered the unit” (Bloom, 1981, p. 171).
He also believes that grades should reflect one’s academic outcomes, and no more. He writes, “It is our position that the grade should reflect achievement of the course objectives. If necessary, effort, conduct, and other social or classroom factors should receive a separate grade but – as far as is humanly possible – should not influence the assignment of the course grade.” This policy sounds very much like modern day standards-based grading advocates, who insist on separating so-called “habits of work” grades from academic skill grades.
Bloom also includes a fairly devastating “take-down” of the 0-100 point scale for grading. It’s all the more impressive because it’s so dry; it’s as though a statistician is auditing you to death. First, he writes, percentage-based tests are not necessarily criterion-referenced; they simply don’t tell you a lot about how you have actually done, or how much you have actually learned. Next, he reminds us that when you’re giving a test with different objectives being assessed, it quickly becomes impossible to reference one student’s score against another’s. One could easily score well on part 1 but not 2, while one’s classmate scores well on the reverse – and yet, both students could earn the same grade. There are more reasons, and they are just as boring, and just as true. Again, it’s hard not to be won over by these arguments, given the scrupulousness with which Bloom, et al make them. This is very different from the familiar sneering tone of the educational consultant, wagging his finger at us, reminding us that our gradebooks “don’t really tell anyone anything.”
As always in Bloom, there’s an admirable focus on clarity. As in his famous Taxonomy, Bloom stresses clarity of objectives, taking us through several approaches to curriculum design, including the famous Tyler approach (from Ralph Tyler) and insisting on defining the goals of learning as not only content to be mastered, but on changes in behavior to be expected. Evaluation is a way to measure these changes in the learner, as opposed to a sorting mechanism to classify learners.
***
Reading this book makes me think much more deeply about my frustrations with modern-day proficiency learning. Bloom makes a powerful case, and his insistence on development over selection is simple, but admirable, and deeply resonant with me. Moreover, his focus is always on two thing that I value deeply as an educator: clarity of expectations, and individualization of instruction. While it’s easy to read Bloom and feel as though one is reading the work of a detached scientist, forgetful of the myriad obstacles from outside of the classroom that affect students’ performance inside of the classroom, and uncensored with the affective dimension of education (Bloom says nothing about relationships or the emotional aspect of teaching), it’s hard not to feel that he, like John Hattie after him, is certainly pointing us toward some of the best of what educators can hope to do.