Showing posts with label attribution. Show all posts
Showing posts with label attribution. Show all posts

Thursday, November 24, 2011

Getting Clear on Outcomes : a preliminary to project planning

There are many examples ... of situations in which disparate groups of politicians and the constituents they represent have joined together in common cause but consensus has represented nothing more than a superficial commitment to a simple slogan. -- Susskind & Cruikshank, Breaking the Impasse, 1987, p.63-64
Suppose we were considering funding a project to improve student scores in mathematics. Suppose it is shown that Johnny Jones has failed a math test. Is Mr. Smith his teacher to blame because he is a bad teacher? Or is Johnny a lackadaisical student? Or was Johnny sick when he took the test? Or something else?

From the meager information that Johnny failed a test, we cannot make a fair judgment. Johnny’s failure might be accounted for by any of the suggested explanations. How could we add information to the skeletal description, "Johnny Jones failed a math test," to make a decision as to what really happened? What information would be critical? What would be non-informative? What our project aims at will depend upon it.

People often jump into planning a project without carefully specifying what outcomes they are looking for. This avoidance is done sometimes because they feel that the commitment to the project is weak and they don’t want to provoke objections. A risky choice! They may finesse agreement in the initial stages of the project only to have crucial support pulled out after vast expenditures of resources. Such has happened, for example, throughout the history of school reform movements in the United States.

Another reason they may avoid careful specification of outcomes is that they think such procedures are a “mere matter of semantics,” a dispensible impediment to “action-oriented” people like themselves to getting things accomplished. This is delusion. It treats superficial consent to vague terms as though it were deep commitment.

If you consider how many things incite calls to action which are later abandoned because hidden disagreements arise to sap them of support, you will likely agree that some kind of process is in order that will help avoid misunderstandings at a basic level. The alternative is to go ahead blindly and hope for the best. From experience we know that the likely result is a vast waste of time, money and effort.

The process we suggest is called case comparison and analysis. In using it we bring up realistic examples for consideration. We note differences and similarities. We decide which cases we, individually, can agree on most clearly represent the critical terms contained in our goal statement. Then we try to spell out what the defining characteristics are for those terms.

Examples of cases for analysis can be found via the following links:
Model Cases: Effective, Moral or Worthy Teaching
Model Cases: Justice, Expedience, Favoritism
Model Cases: Punishment, Cruelty, Infliction
Take a look at the cases involved. You will see that they are very similar. Yet a minor difference in description may involve a major change in how we classify them.

For references and to examine these issues further, see Defeating or Supporting a Case Characterization


Cordially
--- EGR

Friday, November 4, 2011

Will That Dog Bite? Will That Applicant Be an Effective Teacher?

The proof is in the pudding. -- Proverb
A recent (2009) article asks, in its title, an intriguing question: Can You Recognize an Effective Teacher When You Recruit One?. The authors propose testing potential teacher recruits for two factors: “cognitive skills” and “non-cognitive skills.” On the basis of their tests they claim they can predict likely future student academic achievement, thereby reducing or sparing school districts the costs of probationary appointments.

When I applied for the Peace Corps in 1964, I and the other applicants were subjected to two kinds of tests: one for cognitive skills; another, for non-cognitive skills. Those of us who “passed” them nonetheless participated in programs with various outcomes: some successful and some not. I have never run across any indication that program successes or failures showed any correlation with Peace Corp Volunteer scores on either test.

What seems to be going on here with this “predictive testing” is an example of what some psychologists call the Fundamental Attribution Error -- ignoring contextual and environmental factors by locating all causality, all responsibility, in the individual person.

The tests mentioned in the article will likely be very popular with school boards and parents because they basically wipe the slate clean when it comes to asking what influence, and what responsibility, these groups have for student achievement.

For references and to examine these issues further, see Increasing Teaching Efficiency: 
the evaluation of method


Cordially
--- EGR