Conclusions on Models

Ross Woods

Quality is a more complex issue than it appears. Almost every model can be useful at times, and it appears almost impossible to have only one in isolation, let alone set it up as an ideal. All have weaknesses or limitations, and all make assumptions and foreshadow types of program evaluations.

Eclecticism is not a surprising conclusion; the interrelationship between curriculum models largely foreshadows it. If curriculum studies are any guide, it is improbable that accreditors can produce an eclectic model that perfectly balances all constituent models and elements. (Cf. Print, 1987:72) Also like curriculum models, the views of quality implied in varying evaluation models do not always combine into theoretically consistent concepts; some combinations seem all but impossible. A product model at its heart believes in effability so by nature it does not mix with ineffability, or with the extreme versions of the environment-experience model that hold that educators should not predetermine learning outcomes.

Some differences between models are less substantial than they appear, reflecting simply the different emphases, backgrounds, and personalities of different leading evaluators. (Cf. e.g. Stake, 1983:290; Stufflebeam, 1983:122-124)

In practice, accreditors and schools invariably combine different models of quality, uncritically and probably unconsciously. The ATS (1987) mentioned a variety of elements in passing and assumed them all to be valid, including accountability, consensus, felt needs, resources and facilities, professional standards, and learning environment. Hopkins (1987:5) uses the word "quality" to mean accountability, teaching-learning process, and use of resources to create the best possible learning conditions. Solmon (1981) suggested even more aspects in his eclectic view.

The product model answers questions about purpose and tangible results. Interpersonal models are not only useful, they are unavoidable in terms of interest-group ethics. They can produce an abundance of information, even including data on products and objectives. More than that, product and interpersonal evaluation are interdependent because stakeholder evaluation acts like a feedback loop on a means-ends model. When a school reaches the program evaluation stage in the means-ends cycle, it needs feedback from the people in the program (that is, its stakeholders) to understand itself accurately and how to revise its program. Not only that, the cycle also applies to accreditors who need feedback to evaluate their accreditation program.

Even staff holding the most rigid product notions want to come to some kind of consensus on their objectives, and they rely on consensus to interpret them similarly. Schools which hold that quality is largely ineffable always produce literature which describes their programs. In fact, it would be very difficult to find a completely pure example of any one model.

The arguments for responsible eclecticism are very strong. The characteristics of modern program evaluation, its models, and the models of quality tend to harmonize. Accreditation of nontraditional education needs to include most of them.

Eclecticism and the CIPP Model

The CIPP (Context, Input, Processes and Product) model is another major model of program evaluation which can be related to accreditation. As an eclectic model, it is ill-fitted to the previous chapters and well illuminates the role of eclecticism.

The early form of the CIPP model was a solution to some of the weaknesses of Tyler's evaluation model of quantitative evaluation, but it changed over time. In some ways, it still closely resembles Tyler's model of curriculum in that it works from needs to processes and then to fulfillment of need expressed as a product. Unlike Tyler's evaluation model (which used a pretest and a post-test), CIPP could evaluate a program at one time at any stage from planning to implementation.

As its name suggests, CIPP implies that evaluation comprises four loci, which largely reflect the time sequence in developing a program:

Context refers to program justification. This includes defining the target population and their needs and underlying problems, as well as the institutional context.

Input refers to the prescription of a program, especially in terms of alternative strategies, institutional capability, and design practicalities.

Process refers to a check on the implementation of the plan. The process check aims to provide feedback to staff, guide program improvement, assess whether staff roles are appropriate, and document the program.

Product evaluation is the measurement and judgment of the program's achievements and side-effects, especially in terms of whether needs are met. Stufflebeam also mentions input from a wide variety of program participants and the use quantitative evaluation. (Stufflebeam, 1983)

Eclecticism is a strength of the model. It includes many aspects that the omission of which would be a weakness. It includes the key advantages of the product approach, as well as program feedback, program change during implementation, and awareness of the difference between intended and actual results. It also even hints at stakeholderism.

It risks many of the weaknesses in Tyler's early models such as its conceptions of product, and it does little to imply a methodology; it is more like an umbrella that provides a rationale for many techniques and local situations. Although these weaknesses disqualify it as an adequate accreditation model, it does at least show that eclecticism is a viable option.