Siegfried Streufert

Complexity and complex decision making: Convergences between differentiation and intergration approaches to the prediction of task performance ☆

Abstract A series of three experiments designed to measure effects of environmental complexity (information load, failure, and success) and complexity of personality structure on differentiation and intergration in decision making were conducted. It was found that increases in load, success, or failure produced initially increasing and then decreasing intergration levels. Differentiation levels increased and then remained constant. Differentiation exceeded intergration. Structurally complex subjects exceeded structurally simple subjects in intergration scores, but not in differentiation scores. It was concluded that complexity theories dealing with differentiation and those dealing with intergration may not be viewed as equivalent or widely related. Common theoretical predictions for differentiation and integration were placed in doubt.

Perceived control and risk Perceived control and riskiness

Twenty undergraduate students were placed in a complex decision-making situation, involving a simulated internation conflict which approximates real-world decision making in many aspects. The analysis of the Ss’ economic decisions indicated that those Ss who perceived that conditions in the simulated environment were due to their own decisions tended to take fewer risks than Ss who perceived that conditions were due to forces beyond their control.

Success and response rate in complex decision making

Abstract Competing theoretical predictions by Schroder, Driver, and Streufert (1967) , by Streufert (1969) , and by a tentative extension of learning theory for success effects on the components of response rate in complex decision making were tested. Dyad decision-making teams participated in the Tactical and Negotiations Game, a complex experimental simulation technique. Twenty-two dyads were exposed to increasing proportions of success information to neutral information. Eleven control dyads received only neutral information. Since no time-order effects were found, the results could be interpreted as effects of success induction. Increasing success resulted in decreasing integrated decision making, increasing respondent and general unintegrated decision making, and a constant total response rate. Some of these findings were reversed, however, when teams received only success information (i.e., total rather than partial reinforcement). The data were interpreted via a combined effect of Streuferts (1969) modification of complexity theory and learning theory.

Information load, time spent, and risk taking in complex decision making

The effect of the quantity of information which decision makers receive per unit time and the effect of the length of time spent in decision-making groups on the degree of risk taking in decision making were examined. A simulated decision-making task of some complexity was used for data collection to permit comparison with results obtained in simpler laboratory settings. It was found that risk taking increases with time spent in decision making and reaches highest levels under optimal information conditions. Parallels to studies in simpler environments were examined.

The perception of experimentally-induced success

A complex decision-making environment was used to examine the relationship between experimentally-induced success and Ss’ perceptions of success for cognitively simple and cognitively complex Ss. In each case a significant linear relationship between induced success and perceived success was obtained. In addition, higher order trends suggest tentatively that success may constitute a multi-dimensional, rather than a uni-dimensional, phenomenon.

Increasing Failure and Response Rate in Complex Decision Making.

Abstract The effect of increasing failure on response rate and its components in complex decision making was investigated. Data were collected in a simulated decision-making environment permitting integrated, retaliatory, and general unintegrated decision making by dyads. It was found that (1) integrated decision making first increases and then decreases as the failure levels of information which groups experience are increased; (2) general unintegrated decision making increases when failure reaches moderate levels; and (3) retaliatory decision making is not affected by failure induction. The total decision-making response rate generally follows the characteristics of the curve for the integrated decision-making component, although it is, of course, somewhat higher.

INFORMATION SEARCH AND THE EFFECTS OF FAILURE: A TEST OF COMPLEXITY THEORY

Abstract The effects of increasing failure and the structural complexity of decision-making teams on four aspects of information search and utilization were examined. Twenty-four structurally homogeneous dyads and 12 unselected control dyads served as Ss in a complex experimental simulation running for six 30-min periods. Information load was held constant at 7 inputs per period. The proportion of failure inputs was sequentially increased from 1:7 in period 1 through 6:7 in period 6. Dyads in the control condition were not exposed to failure. It was found that structurally simple Ss engaged in more delegated information search than complex Ss. Self-initiated information search initially increased, then remained fairly constant at higher proportions of failure. Contrary to expectations, no differences were found between simple and complex Ss in self-initiated search. Complex dyads exceeded simple dyads on two measures of information utilization. The number of search moves used in integrative decisions produced an inverted U-shaped curve with optimal levels at moderate proportions of failure. Efficiency of information utilization showed a general decrease as failure increased. The data produced limited support for complexity theory and did support, where applicable, the information search theories of Lanzetta and Feather.

Task Familiarity and Reliance on the Environment in Decision Making

The study investigated change in dependency on the environment as a function of increasing familiarity with the situation. 72 students formed 36 dyads which played a complex tactical and negotiations game for 6 one-half hour periods. Over periods, all subjects reduced their use of the environment as a source of influence on their decisions. Cognitively complex subjects reduced their dependency on the environment sooner and more sharply than simple subjects. It was concluded that as a task situation is mastered, decision makers gain the confidence to take a more active role in structuring it. The implications of increased familiarity with a task situation for the shift to risk phenomenon were indicated.

Information Load, Proportion of Relevance, and Relevance Perception.

The effect of experimentally induced relevance on subjective relevance perception was investigated. Ss participated in a complex decision-making task. In Condition A, Ss received 1, 2, 3, or 4 relevant messages out of a total of 10 messages for any one of four (randomized) playing periods. Condition B exposed Ss to 1, 2, 3, or 4 relevant messages out of 20, and Condition C exposed Ss to 4, 8, 12, or 16 relevant messages out of 20. All other messages received by the Ss were irrelevant to their task. Messages had been prerated by a parallel population who did not participate in the complex task. It was found that subjective perceptions of relevance in complex tasks tended to be overestimates of actual relevance levels, particularly when the proportion of relevant information to total information was low. Further, Ss tended to underestimate actual relevance levels when information load of relevant items greatly exceeded the optimum established by complexity theory research.

Effects of failure in a complex decision-making task on perceptions of cost, profit and certainty☆

Abstract Dyad decision-making teams spent seven consecutive half-hour periods playing a complex internation game containing tactical, economic, negotiation and intelligence components. Although the subjects assumed that they were affecting their environment, the game was pre-programmed. Teams received seven informative messages per one-half hour period. Increasing failure was induced by increasing the proportion of failure to neutral messages over periods. After each one-half hour period, subjects indicated their present and anticipated cost, profit and certainty estimates on rating scales. Increasing failure produced generally rising cost level estimates. A discrepancy between actually experienced (present) and anticipated cost levels under low failure levels produced an “optimistic” outlook by subjects that expressed itself in initially steady profit estimates and initially rising certainty estimates. Under higher failure levels, profit and certainty estimates were inversely related to (increasing) cost estimates. It is suggested that under low failure levels, perceptions of psycho-economic conditions in complex environments may be multiply determined. However, as failure levels increase (and optimism is absent), perceptions of psycho-economic conditions may be determined by the most salient environmental characteristic.