Group versus individual problem-solving.

Abstract

Is THE performance of a group superior to the performance of an individual? Studies by Shaw (1932), Watson (1928), andTaylor andFaust (1952) have found that group performance is significantly superior. In the Taylor-Faust study (1952), for example, groups composed of two and four members were found to solve more Twenty Questions problems in fewer questions and in shorter time than did persons working individually.These studies have based their statistical analysis upon the null hypothesis that no difference exists between group and individual performance except that produced by chance. However, the results indicating that group performance is superior to individual performance have often been interpreted as due to some "group effect" in which individual members of the group are influenced by other members to work on problems along lines of thought which they would not have generated themselves. Marquart (1955) has criticized the validity of such an interpretation in which a person working alone is counted as equal to one group and in which no allowance is made for the fact that a group solution might be the result of any one of the members (perhaps the ablest) rather than the cooperative effort of the group as a whole.A different null hypothesis and one more appropriate for testing the influence of some group effects can be derived from the discrete

A Theory of Collective Induction

“…Examples include object transfer problems (e.g., Shaw, 1932), water jar or gold dust problems (e.g., Davis & Fistle, 1963), anagrams (e.g., Faust, 1959), and most of the tasks in the group problem solving traditions summarized by Kelley & Thibaut (1969), Hackman and Morris (1975), Hastie (1986), Hill (1982, Lorge, Fox, Davitz, andBrenner (1958), andMcGrath (1984).…”

Cultural values, educational methods and small group communicator styles in the United States and the People's Republic of China

“…It has been said that the greater a person ts knowl e_dge of a. subject, the 1 ess conformant he or she will be, so perhaps it can be said that the inverse is also true~-the less a personts knowl~dge on a subject, the more conformant he or she will become. This idea is· reinforced by another study which es tab 1 i shed the fact that conformity is increased with i ncre'ased contra l by the agent of conformity (Fearing and Krise 1941). This conformity is self-perpetuating: a study by Scott (1956) showed that the greater the number of members in a_ group who hold a value and the more strongly they hold it, the more important it becomes to all.…”

Creativity in Organizations

“…Faust (1958) identified complex sentence anagram tasks as "multiple-step" tasks, as compared to the "single-step" needed to solve one-word ("Eureka") problems. Groups have been found to perform better at complex tasks, while individuals performed better at the less-structured, single-step tasks (Faust, 1958, p. 17;Lord 1976, p. 87).…”

“…Methodological defects exempted experiments 9 and 10 as the methods for recording responses were different: nominal groups wrote while real groups tape-recorded oral responses. Also exempted was experiment 8 (Faust, 1958) because of the use of complex tasks which required a multiple-step development as compared to the simple single-step responses called for in the other brainstorming studies. After performance percentages were computed for the remaining set of studies (Table 1) Meta-Analysis Rejuvenates.…”