Internal Control Relations in Administrative Hierarchies: A Critique

Abstract

Employing the technique of direct observation, Walter studied the flow of information through the municipal hierarchies of two southern cities. He made observations on two general classes of communicative exchanges: (1) those involving factual and (2) those involving imperative sentences. The former, which Walter divided into sensitizing assertions and problem-solving simply convey information about the task at hand; while the latter, which also convey information of course, are in the form of directives or orders. Problem-solving were subdivided into estimates of costs and estimates of benefits. Thus, Walter presented his findings under three types of factual but under only one category of imperative sentences. Walter further classified communicative exchanges by their reference to either programmed (routine) or nonprogrammed (nonroutine) decision making. He implemented these categories in order to test two hypotheses on interpersonal between formally designated superiors and subordinates. The first hypothesis stated that subordinates would exercise more than superiors in the area of nonprogrammed decisions and the second hypothesis stated that superiors would exercise more than subordinates in the area of programmed decisions. For factual assertions, Walter considered that an instance of effective had occurred if (1) A had transmitted an item of information to B and, (2) B was observed to behave in some way which indicated that he had accepted the information as a basis upon which to predicate his own actions. Given such an occurrence, A had made an attempt and the attempt had been successful; i.e., A had achieved an instance of effective on B. For imperative sentences, Walter considered that an instance of effective had occurred if (1) A had transmitted an imperative to B and, (2) B was observed to behave in accordance with the imperative. Given such an occurrence, A had made an attempt and the attempt had been successful; i.e., A had achieved an instance of effective on B. After collecting data on a rather large number of communicative exchanges (apparently more than 32,000) between superiors and subordinates, Walter (1) took a ten percent sample of his observations from each city, (2) grouped these samples, and, within this combined sample, (3) lumped all superior-initiated exchanges together in one subcategory and all subordinate-initiated exchanges together in another. He then proceeded to test his hypothesis for each type of communicative exchange with a measure of interpersonal which he called the influence index. In terms of this measure, Walter claims to have confirmed both hypotheses for the data on imperatives, but only the hypothesis on programmed decisions for the data on factual assertions.