Peter Krope, Peter Petersen

1

H_S_ENG1.DOC

Peter Krope, Peter Petersen

Generic Acts. The Constructivistic Contribution to the Solution of the Body-and-Mind Problem

Prof. Dr. Peter Krope
Dipl.-Päd. Peter Petersen
Zentrum für Konstruktive Erziehungswissenschaft
Institut für Pädagogik
Universität Kiel
Olshausenstraße 75
D-24118 Kiel
Tel. (0431) 8801273
Fax (0431) 8801588
Email:
Extended version of a Paper presented at the European Conference on Educational Research, University of Hamburg, 17-20 September 2003

1The Problem

The background of this presentation is the research project Report2000. The subject of the actual section is the modern commented educational report (Berichtszeugnis, Krope et al. 2000). It is the type of report, which has been introduced frequently in the FRG as an alternative to the traditional mark report. The new type may be regarded as extremely controversial. The attitudes range from strict disapproval to decided agreement. The research project is based on the assumption that the differences may be explained by a fundamental disagreement on the goals which such a report is supposed to serve. Within this discussion, a rational dialogic method is tested, for a non-dogmatic, circle-free agreement on incompatible goals. It is intended to control the result of this intervention by measuring changes in attitude using objective methods.

While developing a method to measure attitudes, a problem has arisen which is the subject of this presentation. Someone, who intends to produce empirically true statements about persons, for example about their opinion concerning the commented educational report, has to rely on a theory for the application or development of tests and questionnaires. At present, two relevant test theories are available. They are the classical and the probabilistic test theories. The problem of this presentation – the central issue of this paper - could principally be demonstrated by both of them. For its clearer representation, the probabilistic test theory is chosen here. The general starting-point is: P(avi) = f (V1, V2, ...). The formula means that the probability P of a special reaction a of the person v to the item i is a function of the latent variables V1, V2, ... (ROST and SPADA 1978, p. 63/64). In the formula something manifest is mentioned indirectly, that is to say the reaction of a person. In the context of multiple-choice items, it is the behaviour of choice in ticking boxes. In the formula something latent is directly mentioned which is essentially identical with terms of disposition. A simplified explanation suggests that the probability of the (manifest) behaviour is explained by (latent) dispositions (ROST 1995). Theoretically, we may assume that the probability of manifest facts may be explained entirely by identifying a finite number of latent variables. This is based on models, which axiomatically describe the connection between the manifest and the latent parameters (SCHEUCH and ZEHNPFENNIG 1974, p. 134).

An axiomatic description is a problematical base for methods intended to give information about persons. In fact, axiomatic theories have no empirical relevance at all in so far as their basic symbols are not semantically interpreted (BUNGE 1967). But if the elements of an axiomatic system cannot be interpreted according to rules, the necessary condition for the validation of empirical statements about persons is no longer applicable. Due to this restriction another problem arises. In analysing the latent structures, the probabilistic test theory may overcome the instructions of extreme behaviourism, which prescribes the observation of external behaviour as opposed to internal motives (WATSON 1966). These models connect external with internal dimensions of the tested persons. But the answers to the question regarding the type of relation between the manifest and the latent dimension are nothing but correctly formed logical sentences in an axiomatic theory. The question of their validity may only be answered within the restriction of the mathematical description of the test method. This test theory which divides its subject into two entities creates a problem which has a parallel in the body-and-mind problem of philosophical tradition. The attempts to solve this problem which are caused by the search for the relation between body and mind cannot be proven empirically (POPPER and ECCLES 1982).

The fundamental objections against the classical and the probabilistic test theory have been the reason for developing a constructivistic theory of educational measurement (KROPE 2000a). Its basis is the Methodical Constructivism (the Constructivism of the Erlangen School; compare KAMLAH and LORENZEN 1973, LORENZEN 1987). For the purpose of the research project Report2000 the constructivistic theory has been developed further, an important part of which is the principle of generic acts (Handlungsschemata). It is to be expected that this principle combined with the constructivistic philosophy presented by HARTMANN (1998), will contribute to a solution of the body-and-mind problem, which is of importance beyond the research project Report2000.

2Introducing the Terminology

Within the research project Report2000 the attitudes to the commented educational report are to be measured. While the terms „to measure“ and „commented educational report“ have already been introduced elsewhere (KROPE 2000a, 2000b) we are trying in the following to define precisely the term „attitude“ as it is called in every day language. For this purpose we are starting from the early beginnings of the living world (Lebenswelt) according to the constructive science program. Based on this pragmatic level, the elements of scientific language are being introduced according to the rules of methodological thinking, in a comprehensible, sequenced manner and without any preconditions.

The fact that advice is frequently sought on the commented educational report is regarded as an expression of an interrupted continuum of action within the everyday school situation. The commented educational report meets approval on one side, disapproval on the other. The controversy is reflected in the jurisdiction. In Schleswig-Holstein digit based marks have been forbidden to a certain degree and then permitted again within the last decade. The practice of issuing reports at schools has not been conflict free lately.

For the intervention in this conflict, lectures, discussions and trainings are offered within the framework of the project Report2000. A sample report has been developed for these events (KROPE 2000b) which may be used as illustration, as instruction for the creation of a commented educational report and for the explanation of its principals. The sample offers a variety of options. The basis of the various interventions is the rational dialogue (KLARE and KROPE 1977). It assumes that for the opening of a dialogue one debate partner makes a suggestion, which is being accepted, rejected or substituted by other proposals until an agreement has been reached.

For the debate according to the pattern of the rational dialogue, one member of the research team acts as proponent. In the debate, the proponent offers his proposals with sentences such as „Here, this is my suggestion“ and illustrates the advantages of the individual options. Since the opponents criticise different facts, the proponent is demonstrating the various positive options of his proposal, which are designed to combat the disadvantages. In his challenges, the proponent uses various remarks as answers to different reproaches. He may say: „Accept it!” or „Adopt it!” If the opponents follow the suggestion of the proponent, we may say that they are in agreement. We stipulate to call all synonyms for such remarks of the proponents, which have been related to the opponents in first person, „agreement“. If the opponents do not agree, the respective remarks will be called „doubts”. Basis for the consideration of the synonyms was DUDEN, Volume 8 (WISSENSCHAFTLICHER RAT DER DUDENREDAKTION 1972).

According to this convention „agreement“ exists when the opponents declare after a dialogic debate on the commented educational report: „I approve of it”, „I find it remarkable”, „I am all for it”, „I appreciate it”, „I accept it”, „I welcome it”, „I think it is desirable”, „I accept it”.

Accordingly, we are speaking of „doubts“ when the opponents observe about the commented educational report the following: „I am against it”, „I reject it”, „I disapprove of it”, „I criticise it”, „I have doubts about it”, „I protest against it”, „I find it untenable”, „I do not agree on it” (A similar introduction of the terminology „agreement” and „doubt” is found in GETHMANN 1978, pp. 143 – 145).

We are presuming that the opponents relate to different states of affairs (Sachverhalte) mentioned in the dialogue when they choose different phrasing in order to express their agreement or doubt.

3The Predication on Empirical Base

In the last chapter the terms „agreement“ and „doubt“ have been introduced. In the following a method will be explained which can measure states of affairs such as agreement and doubt. The method is called predication on empirical base (KROPE 2000a). Like the constructive use of terms, the constructive measure method differs from the traditional approach of the probabilistic test theory. Like for the above terms, the theoretical base of the measure method is being displayed.

KAMLAH and LORENZEN (1973) and LORENZEN (1987, p. 25f) begin the construction of language by practising words through modelling and copying. Once a word like „to agree“ has been practised and it may be assumed that MR A. has agreed to the commented educational report, you can point to MR A. and say: „This man has agreed to the commented educational report”. This way a word is assigned to a person. This process is called predication, the word which is assigned to the person, is called predicator.

For this example the question may arise how MR A. might behave in reality. This question aims to clarify whether MR A. has really agreed upon the commented educational report, since - for this example - the agreement has only been assumed. Because for constructivism the term „real“ is defined with the help of the term „true“ and not the other way round, further differentiation is determined. Someone might point at MR A. and comment that MR A. has agreed. Then, someone else likewise pointing at MR A. responds that MR A. has not agreed. Sentences, which can be claimed or rejected in this manner, are called statements. Statements are regarded as true within constructivism as far as rules exist, according to which these can be defended against any objection.

Elementary statements like „Kiel is a city“ are called empirically true statements. Empirically true statements can be made about single objects (Gegenstände), events and persons. Those who pretend to formulate empirically true statements about persons and who refuse to demand non-sustained credibility are faced with the task to find out information about these persons via empirical procedures. An experiment is one of the empirical methods. The requirements of empirical methods are described as follows for an experiment. Pointing towards the experimental natural sciences JANICH (1996) describes four characteristics of an experiment, which are also valid for surveys on persons according to the constructive viewpoint. First of all you have to act poietically in an experiment, secondly, experiments have to be repeated and varied, thirdly, every experiment has theoretical requirements and fourthly, the data of an experiment need to be combined for a result (JANICH 1996, pp. 277-279). In connection with the considerations at issue, we are looking at the first characteristic, which is the poietical acting. JANICH calls it the manual-technical side of the experiment and distinguishes three phases: the preparation, the start and the registration. He uses the classical fall studies as an illustration, which Galileo GALILEI carried out with a ball, an inclined plane and a water-clock. The expression „preparation“ relates to the structure of an experiment, „start“ refers to the release of the ball, and „registration“ means the data on the way of the rolling ball in relation to time. The start of the experiment is the link between a scientist gaining insight and reality. „The 'start' marks exactly that event which becomes relevant for an experiment of the natural sciences. It is like an interface between a state of affairs created by human acts – hence by using the hands – and a process which neither represents a human act nor is disturbed by one“ (JANICH 1996, p. 278). The scientist, who pretends to be in possession of knowledge, formulates statements about his claimed knowledge – GALILEI about the ball, the educational scientist about persons. After the start of the experiment these statements are subject to the risk of failure.

For the preparation of the experiment in the research project Report2000 a questionnaire has been developed for the measurement of agreement and doubt. For that purpose the eight remarks mentioned above which each represent either agreement or doubt have been chosen from a large number of synonyms and used as a base for the development of the questionnaire. The three criteria for this choice have been firstly ease of understanding. Even persons with difficulties in reading should have been able to understand the items. The second criterion was the economy of the experiment. The questionnaire has been designed to require a limited period of time to be processed. The third criterion was called „differentiation“. It should be possible to determine an increase as well as a decline in the number of remarks in agreement or in doubt subsequent to an intervention. The chosen two times eight synonymous remarks have been included as items in the questionnaire. The questionnaire serves to answer one single question only. It is directed at the person being interviewed following the instructions and is phrased: „What do you think about the commented educational report?" The questions may be ticked as „yes“ or „no“ for an answer. The first two items are:

YES / NO
I appreciate it...... /  / 
I disapprove of it...... /  / 

The questionnaire relates to a debate where agreement or doubt may be expressed during the process and in changing situations. Hence, you may count for the evaluation in how many situations a person has agreed or doubted, and indicate the result in points from 0 to 8 as agreement or doubt. The result is a ratio scale.

The empirical survey which is described in the following, starts with the problem which stems from the fact that the test persons have not participated in the agreements on terminology of the research team. Referring to the described development of the questionnaire it was assumed that these persons followed the agreements on terminology for the survey „as if they had been present”. This assumption was validated as follows. The hypothesis for the empirical study was derived from a course in advanced studies on the commented educational report, which had been arranged for elementary school teachers. The aim of the course was to illustrate that writing commented educational reports requires extensive knowledge of theory and methodology. According to the recorded observations most of the teachers have been discouraged by the high requirements but not motivated to participate in a following course. While in the beginning about 95% of the remarks provoked by Metaplan method were agreements to the commented educational report, in the final discussion exclusively doubts have been contributed. On the basis of these observations the following hypothesis for the empirical study was formulated: After a course arranged highly towards theories and methods the test persons express less agreement and more doubt for the commented educational report than before. The hypothesis was validated in two groups of persons. For group 1 a course arranged highly towards theories and methods was carried out on the commented educational report. Before and after the course agreement and doubt were measured using the questionnaire. This group may be called test group. For group 2 the measurement of agreement and doubt was being repeated within a period of two hours as well, whereby an instruction on the commented educational report did not occur in the meantime. This group may be called control group. The structure of the experiment is designed as pre-test and post-test with test group and control group. Since the two groups did not represent real samples and neither random choice nor a randomisation were intended, this method is called a pre-experimental design according to the system conceived by CAMPBELL and STANLEY (SCHWARZ 1970).

pre-test / post-test
agreement / doubt / agreement / doubt
/
s
n / 6.05
2.99
21 / 1.95
1.80
21 / 4.82*
3.19
17 / 3.18**
2.51
17
/
s
n / 5.9
3.1
78 / 1.99
2.8
78 / 5.9
3.1
79 / 2.1
2.8
79

Tab.: Interim results of the research project Report2000, release Spring 1999. The persons of the test group are teacher students preparing their final examination, the persons of the control group are students who study for a Diploma degree in education at Kiel University, Seminar Statistics II; identical experimenter for test group and control group. Keys: - arithmetic mean, s – standard deviation, n - number of test persons. The marked post-test-values differ significantly from the corresponding pre-test-values (* -  = 15%; ** -  = 5%; one-tailed test).

The results from the current survey are displayed in the table. As predicted in the hypothesis, the number of remarks in agreement has really declined for the test group whereas the number of doubtful remarks has risen. The differences are significant concerning the doubt ( = 5%) and concerning the agreement ( = 15%, computed for heuristic purposes). A first interpretation of this result considers the fact that the persons questioned in the study were not informed about the agreement on terminology of the research team. Therefore the danger exists that interviewers and interviewees talk past each other. At the same time, the design with pre-test and post-test and test group and control group precludes the assumption to a certain degree that the surveyed persons would express anything else but agreement and doubt for the commented educational report – such as sympathy for the experimenter – within the questionnaire.

For the further interpretation of the result we fall back again on the constructive theory. According to a distinction made further above, real states of affairs are expressed by empirically true statements. Therefore, we may continue to draw the following conclusion from the results in the table. Within the limits determined by the pre-experimental design, we may say that the persons of the test group have really less agreed and really increasingly doubted. By assigning the predicators „less agreement“ and „more doubt“ to the persons of the test group, a statement is made on reality. With the expression „reality“ we are relating here once again to the living world from where the study Report2000 has started with the formulation of the problem and the creation of the terminology.