Chapter 15 page 359
Response. The lesson plan shares many elements with STAD. Students work in groups.
They are given a group reward, and the group reward is based on average student
performance. However, there is also a crucial difference between this activity and
STAD. In the STAD procedure described in the test, the reward was given for average
individual learning on a test. There was no way for one student to help another on the
test; therefore, the test assessed what each individual had learned. In Ms. Amborn’s
plan, the individual work is an advertisement, not a test, and students can help each
other freely on it. If one student—say, Eric—does not understand the propaganda
techniques, there is no need for other students to help him understand. They only have
to tell Eric explicitly what to put in his ad. To make sure that Eric gets a high score on
his ad, the most proficient student in the group might even do Eric’s advertisement for
him, leaving only the artwork for him to do. Eric can get a high score on the
advertisement just by doing what his peers tell him to do, even if he does not
understand the propaganda techniques at all. For this reason, Ms. Amborn’s plan falls
short of being STAD. It creates some interdependence, but it does not require students
actually to help each other learn and understand the material.The individual reward structure also presents obstacles to effective collaborative learning. For
example, there is no incentive for proficient students to help a lower-performing peer in the group if the
proficient students don’t stand to benefit. Consider a case in which a teacher asks students to study spelling
words together in groups. If the most skilled speller in the group knows that she can get 100% on the
spelling test on her own, she may feel no desire to help another student in the course.
Research evidence on reward structures. In reviews of the literature on cooperative learning,
Slavin (1983; 1984) examined experiments contrasting cooperative learning and individual
(noncooperative) learning in regular classrooms. He found that cooperative learning was clearly superior to
individual learning when there were group rewards for individual learning. When this reward structure was
incorporated into cooperative learning, 24 of 27 studies (89%) showed positive effects of cooperative
learning over individual learning. Among studies that used group study and individual rewards, 63%
showed positive effects for cooperative learning. Among studies that used group rewards for group
products as the reward structure, only 38% showed positive effects for cooperative learning.
Despite these positive results for group rewards for individual learning, there are three additional
points to keep in mind about STAD and related reward-based methods.
Studies that support group rewards for individual products have not measured the role of intrinsic
motivation as a possible outcome. Thus, it is possible that academic gains occur at a partial cost to
intrinsic motivation.
For many educators, one goal of cooperative learning is to promote genuinely prosocial, caring,
altruistic behavior by students. However, students are not—by definition—learning to help others
altruistically when they are always rewarded for any help that they give (Batson, 1991; Noddings,
2002).
The tasks used in the studies cited in Slavin’s classic review tend to employ fairly low-level cognitive
learning tasks, typically involving reviewing material on worksheets provided by the teacher. As we
will discuss in a later section, many collaborative learning experts urge teachers to use complex
tasks at a higher cognitive level. A reward structure that is successful with lower-level cognitive
tasks may be unnecessary for higher-level tasks with much greater situational interest, such as
working on complex projects with topics chosen by students (Cohen, 1994b).
These three points suggest that teachers may want to consider alternatives to STAD, especially when they
are engaging students in more complex projects with high situational interest.