European Review of Social Psychology

ISSN: 1046-3283 (Print) 1479-277X (Online) Journal homepage: https://www.tandfonline.com/loi/pers20

Paradoxical Effects of Praise and Criticism on

Perceived Ability

Wulf-Uwe Meyer

To cite this article: Wulf-Uwe Meyer (1992) Paradoxical Effects of Praise and Criticism
on Perceived Ability, European Review of Social Psychology, 3:1, 259-283, DOI:
10.1080/14792779243000087

To link to this article: https://doi.org/10.1080/14792779243000087

Published online: 04 Mar 2011.

Submit your article to this journal

Article views: 820

View related articles

Citing articles: 5 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=pers20
 Chapter 9

Paradoxical Effects of Praise and

Criticism on Perceived Ability
Wulf-Uwe Meyer
Universitat Bielefeld

ABSTRACT
In this chapter the effects of praise and criticism are considered from an
attributional perspective. It is shown that-when analyzed from this
perspective-praise and criticism may have consequences that seem nothing
less than paradoxical as compared to a reinforcement standpoint, i.e. the point
of view that ascribes to positive and negative verbal feedback corresponding
positive and negative behavioral, emotional, and motivational effects. Empirical
evidence presented shows that praise can lead to the inference that the other
person evaluates the recipient’s ability as low, while criticism can lead to the
inference that the recipient’s ability was estimated as high. It is also shown
that evaluative feedback can have paradoxical effects on the recipient’s self-
perception of ability. Finally, some implications of the research are spelled
out for the social psychology of the classroom and for the development of the
self-concept.

Praise and criticism are social events that are frequently analyzed from the
perspective of reinforcement. Both are then assumed to affect behavior in a
specific way when made contingent upon performance and when they are
specific and appear sincere (O’Leary & O’Leary,1972). Praise is supposed to
strengthen the probability of particular behavior; criticism, on the other hand,
~~

European Review of Social Psychology, Volume 3 Edited by Wolfgang Stroebe and Miles Hewstone
@ 1992 John Wiley & Sons Ltd
260 WULF-UWE MEYER

is assumed to decrease the probability of behavior. Because praise and criticism

are readily at one’s disposal, and because they do not have the disadvantages
associated with material reinforcers, both are widely recommended for use by
teachers (for a review see Brophy, 1981) and parents. Praise, especially, is
frequently considered to be a desirable form of positive reinforcement, not
only because positive behavioral effects are said to stem from praise, but also
because it is believed that praise leads to positive affect in the recipient, thereby
increasing self-esteem. Gage and Berliner (1975, p. 703), for example, contend
(although later qualifying their statement somewhat):
Praise needs no clarification. When the teacher says, “Good,” “Correct,”
“That’s fine,’’ and many equivalent things, we can assume that the message is
intended and received as a positive reinforcer. According to operant-
conditioning theory, such reinforcers should strengthen whatever student
behavior they follow.

There is no doubt that praise and criticism can have the described effects.
However, they do not necessarily always have them. Praise and criticism
can be interpreted in many different ways, and these interpretations then
determine how the recipient responds to being praised or criticized (Brophy,
1981; Kanouse, Gumpert & Canavan-Gumpert, 1981; Meyer, 1978). The
effects of praise and criticism are therefore not straightforward and invariant,
but are mediated by the recipient’s processing of these events and thus can
be manifold. The analysis of praise and criticism solely from the perspective
of reinforcement is far too simplistic.
In this chapter, I will consider praise and criticism from an attributional
perspective. I will show that-when analyzed from this perspective-praise
and criticism may have consequences that seem nothing less than paradoxical
as compared to a reinforcement standpoint, i.e. the point of view that
ascribes to positive and negative verbal feedback corresponding positive and
negative behavioral, emotional, and motivational effects. I will present
empirical evidence documenting that praise can lead the recipient to infer
that the other person evaluates his or her ability as low, while criticism can
lead the recipient to conclude that his or her ability was estimated as high.
I will also show that praise and criticism may affect the recipient’s self-
perception of ability. At the end of this chapter, I will spell out some
implications of the research for the social psychology of the classroom and
for the development of the self-concept.
Let us consider the inferences from praise and criticism and the effects of
these inferences on self-perception and psychological functioning more closely,
beginning with very simple examples. Imagine that a teacher praises you
extensively for success at a task that almost everyone could carry out. You
might conclude from this reaction that the teacher thinks you are not very
gifted. Or imagine that you and another person give correct answers to eight
PARADOXICAL EFFECTS OF PRAISE A N D CRITICISM 261

out of ten tasks. You are criticized by a teacher for this result, whereas the
other person is praised by the teacher for the identical performance. These
reactions may lead you to assume that the teacher views your ability as high,
but views the other’s ability as low. If your self-perception of ability is affected
by the inferred opinion of the teacher, then praise may have ‘negative’
consequences and criticism may have ‘positive’ consequences. Praise may
lead to an estimate of low competence, to dejection or similar negative
emotions, and may bring you to abandon this type of task. A reprimand,
on the other hand, might lead you to conclude that your capability is high
and could give rise to heightened expectations of success, which may in turn
result in increased persistence and performance intensity.

PSYCHOLOGICAL PROCESSES UNDERLYING

THE EFFECTS
The psychological processes and the temporal order of events underlying
these presumptions are shown in Figure 9.1. It is assumed that (1) assessments
of an actor’s ability and effort (2) affect the evaluative reactions (praise and
criticism) toward the actor’s performance. These reactions can (3) provide
the actor with information about how the evaluator is estimating his or her
ability and effort. Under particular circumstances, these inferred opinions
of the evaluator will (4) influence the actor’s self-perception of ability, which
is one of the factors that ( 5 ) determines his or her expectations of success,
affective reactions, and performance.
Elsewhere, I have argued that beyond evaluative reactions other seemingly
positive behaviors may function quite unintentionally as low-ability cues
(Meyer, 1982, 1984). These behaviors include unsolicited help, communicat-
ing pity following failure, and the assignment of easy tasks. Furthermore,
behaviors that are frequently considered to be negative may function as
high-ability cues: withholding of help, communicating anger following failure,
and the assignment of difficult tasks. The present analysis, however, is
mainly confined to seemingly paradoxical effects of praise and criticism: I
have called these effects seemingly paradoxical because they can be
understood from an attributional perspective.

Assassmmls of an Evilluatrue R e a * ~ n s Actor’s Inference ActW’S ExpectatKns.

AUor’s AbWy + Toward the Actor’s the Evaluator’s 4 W - p e r c e p t m -b Emohom. and
and Effort perfamanw As- (Abw 01 AJnlny manor
end E m )
(1) (2) (3) (4) (5)

Figure 9.1 Psychological processes and temporal order of events

262 WULF-UWE MEYER

How does the recipient of the feedback come to know that praise for
success, or criticism for failure, may indicate that his or her ability is
perceived to be low or high? I have assumed that, in adults, these inferences
are based on two attributional principles (Meyer, 1978; Meyer er al., 1979).
First, there is a relation between praise for success and criticism for failure
and the perceived causes of these outcomes. Many studies have shown that
the amount of effort attributed to an actor by an evaluator is one of the
major determinants of the evaluator’s praising the actor for success and
criticizing the actor for failure (e.g. Lanzetta & Hannah, 1969; Weiner &
Kukla, 1970): praise and criticism are maximized when an outcome is
ascribed to high or low effort, respectively (for a summary, see Weiner,
1986). Thus, the evaluative principle refers to implicit knowledge about the
relationship between effort as a perceived cause of success and failure on
the one hand, and feedback for success and failure on the other.
Second, the compensatory principle refers to implicit knowledge about
the degree of effort that4epending on the amount of ability-is necessary
or sufficient for success on tasks of varying difficulty (see Heider, 1958;
Kukla, 1972; Meyer, 1973). In tasks that are normatively very easy (defined
by the performance of a comparison group), ability and effort are perceived
as being compensatory. Given low ability, increasingly high effort is regarded
to be necessary for success on such tasks, until eventually a point is reached
at which ability is so low that even maximum effort is considered to be
ineffectual. In normatively difficult tasks, in contrast, given high ability,
high effort is regarded to be necessary for success. Low ability cannot (or
only within narrow limits) be compensated in these tasks, so that increased
effort is considered to be ineffectual.
Thus, if feedback permits the actor (or a third person observing the
interaction between evaluator and actor) an effort-inference on the basis of
the evaluative principle, then further inferences on the assessment of an
actor’s ability are possible on the basis of the compensatory principle. This
deductive process can be described formafly as a psychological syllogism
(see Figure 9.2; older subjects: compensatory principle. The Halo schema
shown in Figure 9.2 will be described later). Praise for success at an easy
task, for example, can lead to the conclusion that the provider of feedback
attributes the result to high effort. And according to the second principle,
further inferences then can be drawn about how the actor’s ability is
estimated: success is particularly assigned to high effort in easy tasks when
the actor is considered to be low in ability. Praise, therefore, may function
as a cue for inferring that the actor’s ability is estimated to be low.
Criticism can also have such an informative function. Negative feedback
for failure on a difficult task can lead the actor (or a third person) to
conclude that the result was attributed by the evaluator to lack of effort.
However, low effort is the perceived cause of failure in difficult tasks when
PARADOXICAL EFFECTS OF PRAISE AND CRITICISM 263
Praise/Criticism

I Evaluative Principle I 1. Premise

1
Older subjects:
Compensatory principle
2. Premise
Younger subjects:
Halo schema

Ability inference Conclusion

Figure 9.2 Ability inference as a psychological syllogism

the actor’s ability is considered to be high, inasmuch as very low ability

would lead to failure, no matter how much effort was expended. Criticism,
therefore, may function as a cue for inferring that the actor’s ability is
estimated to be high.
These predictions have been tested in a series of experiments (for a
summary see Meyer, 1984; Meyer, Bedau & Engler, 1988; Lebn-Villagra,
Meyer & Engler, 1990). The research varies in that some of the studies
used teachers as the subjects who made judgments. Other studies analyzed
the judgments of students of varying ages. In some experiments ability
estimates were assessed directly, while in others more indirect indicators of
ability were obtained, such as predictions of success. Subjects in some
studies were asked to evaluate fictitious questionnaire situations, and in
others praise and criticism were induced in the laboratory. In spite of these
variations, the general pattern of data was quite similar for adult subjects.
Furthermore, authors from other laboratories have replicated these findings
(e.g. Barker & Graham, 1987; Blickle, 1990; Blickle & Groeben, 1988;
Fischer, 1982; Graham, 1990; Groeben & Blickle, 1988; Hofer et a f . , 1982;
Hom, 1991; see also the special issue “Paradoxical effects of praise and
criticism” of the Zeitschriff fiir Piidagogische Psychologie, 1988 2(4), edited
by Rheinberg).
Some authors have challenged the attributional interpretation of the
seemingly paradoxical effects of praise and criticism. Hofer ef af. (1982) and
Hofer (1985) have proposed that feedback informs the recipient about the
264 WULF-UWE MEYER

evaluator’s performance expectations. Praise indicates that the performance

was above expectation; criticism indicates that performance was below
expectation. Praise, therefore, may lead to the inference that the evaluator’s
expectation regarding the recipient’s performance was relatively low, while
criticism may lead to the inference that the evaluator’s expectation was
relatively high. Thus, inferred unexpectedness of performance rather than
the two principles mentioned above may provide an explanation of the
paradoxical effects of praise and criticism.
This interpretation implies that the evaluator’s high or low ability estimates
can be directly inferred from his or her high or low performance expectations.
However, expectations regarding an actor’s performance are not only
determined by estimates of the actor’s ability but also by the expected or
assumed effort. For example, a teacher may have equally high performance
expectations for two pupils at an easy task, because the teacher thinks that
one of the two students has low ability and will expend very high effort,
whereas the other student has high ability and will expend low effort. Thus,
there is no perfect correspondence between expected performance and
perceived ability. Therefore, performance expectations seem to be no
appropriate basis for inferring ability estimates.
Blickle and Groeben have criticized the attributional interpretation from
the perspective of speech-act theory (Blickle, 1990; Blickle & Groeben,
1988, 1990; Groeben & Blickle, 1988). They proposed an alternative model
for explaining the paradoxical effects of praise and criticism. Reisenzein
(1990a,b), however, has challenged the Groeben and Blickle critique.
Limitations of space do not allow me to discuss the controversy here. Let
me just point to the fact that the Groeben and Blickle model applies only
when the recipient of feedback attributes to the evaluator the intention to
convey his or her ability estimate through feedback. However, it seems
unlikely that the recipient imputes such an intention to the evaluator (see
Reisenzein, 1990b). A student, for example, who is praised by a teacher
(while another student receives no praise for the identical performance) will
hardly impute to the teacher the intention to communicate an estimate of
low ability via praise.

INFERRED OTHER-PERCEPTION OF ABILITY

In the first studies we examined the extent to which praise and criticism led
the subjects to conclusions about the evaluator’s judgment of the actor’s
ability. Effects on the actor’s self-perception of ability were the object of
later investigations. In a study by Meyer et af. (1979, Experiment 3), student
subjects evaluated four versions of a short questionnaire describing two
children who were assigned either a very easy or a very difficult math
PARADOXICAL EFFECTS OF PRAISE A N D CRITICISM 265

problem. Both either succeeded or failed at the task. The evaluative feedback
given by the teacher was different for the two students, although their
performances were identical. Given success, the teacher told one of the
children that the answer was correct (neutral). But the teacher praised the
other child, saying, “You have done very well. Excellent!” Similarly, given
failure, the teacher simply told one child that the answer was incorrect
(neutral), and in contrast, the teacher criticized the other child for the
incorrect response saying “What have you done there! Thirty-five is not
correct.” Subjects indicated on a 9-point scale how the teacher had assessed
the ability of the two children. The four situations (questionnaires)
represented the combination of two outcome levels (success, failure) with
two task difficulty levels (very easy, very difficult).
The mean ability ratings are shown in Table 9.1. As indicated in this
table, the student who was praised after success is rated lower in ability
than the student who received neutral feedback. This is true for both levels
of task difficulty. In the failure condition, the student with neutral feedback
is rated lower in ability than the student who was criticized. This effect is
also present at both levels of task difficulty. Thus, differential feedback for
identical performances may provide information about how an evaluator
estimates the actor’s ability. Praise for success and neutral reactions to
failure can lead to the conclusion that the actor’s ability is considered to be
low. On the other hand, neutral reactions after success and criticism for
failure can cue off the inference that the actor’s ability is perceived as high.
In the experiment just described, subjects were explicitly asked to rate a
teacher’s ability estimate regarding two students. On the basis of this reactive
procedure, however, it is not clear whether subjects would also draw
spontaneous ability inferences from praise and criticism when ability is not
brought to the subjects’ attention via rating scales. Therefore, instead of
introducing ability scales, Meyer et al. (1979, Experiment 6 ; 1988) simply
asked the subjects to give the reason for a teacher’s differential feedback
with respect to the success or failure of two students. Analysis of these free
responses revealed that 70% (Meyer et af., 1979) and 83% (Meyer et al.,
1988) of the subjects attributed the differential reactions to differences in

Table 9.1 Mean ability ratings

Task difficulty Success Failure

Neutral Praise Neutral Criticism
Easy 5.53 2.59 3.18 5.65
Difficult 6.65 4.41 4.77 6.24

From Meyer er al., 1979. p. 264. Copyright 1979 by the American Psychological Association,
Inc. Reproduced by permission.
266 WULF-UWE MEYER

the teacher’s estimate of ability, intelligence, or performance level regarding

the two students. These high percentages indicate that praise and criticism
can lead to conclusions about the evaluator’s ability estimates, even when
ability is not indicated as a possible cause of the differential feedback.

THE MEDIATING ROLE OF EFFORT

INFERENCES
The inferential process from feedback to perceived ability is hypothesized
to be based on the evaluative principle: praise and criticism are assumed to
contain implicit information about effort ascribed to the actor. Inferences
regarding effort then play a central role in forming conclusions about ability
(see Figure 9.2). If the notion regarding effort is correct, then explicir
information about the actor’s effort, which is given in addition to the
information about feedback, should have systematic effects on the inference
of ability.
Meyer, Ploger and Conty (1980; also reported in Meyer, 1984) explored
this possibility in a questionnaire study. Student subjects received one of
four versions of a short questionnaire. Each version described two boys who
had solved an easy math problem. One child was praised by the teacher
(“Excellent! You did it very well”). The other child was told only that his
answer was correct (“Yes, 32 is correct”). Subjects were asked to rate the
teacher’s estimate of each child’s ability on a 9-point scale.
The four questionnaires differed in terms of how much effort information
was added to the teacher feedback information. In a first version (control
condition) no information about effort of the two children was given. In
this condition, the praised student was expected to be rated lower in ability
than the student who received a neutral reaction. In the second version,
subjects were informed that both students-in the teacher’s eyes-exerted
high effort in solving the task. This information was expected to decrease
or to obliterate the differences in the ability ratings between the two
students, relative to the control condition.
In the third version, subjects were informed that the praised student-
from the teacher’s perspective-expended high effort in solving the task. In
this condition the ability ratings for the two students were expected to be
approximately the same as those in the control condition. This expectation
was based on the fact that the explicit information about high effort agreed
with the effort information that-in the control condition-is implicitly
contained in praise. Finally, in the fourth version, subjects learned that the
student receiving a neutral response exerted high effort from the teacher’s
standpoint. The explicit effort information in this condition contradicts the
information on effort that is otherwise implicitly contained in feedback. We
PARADOXICAL EFFECTS OF PRAISE A N D CRITICISM 267

expected that in this condition the ability ratings for the two children would
be the reverse of those in the control condition: the student receiving neutral
feedback was assumed to be rated as relatively low in ability, based on the
teacher’s belief that this student had to exert high effort to solve the easy
task. The praised student was expected to be rated as relatively high in
ability, because in light of the instructions the student would have been
perceived as exerting less effort; the teacher’s praise of this student was
expected to be attributed to factors other than effort.
The results of the study are summarized in Figure 9.3. Ability ratings
from the control condition are represented by the dotted lines in each of
the three parts of the figure. First, the results from the control condition
replicate our previous findings perfectly: the praised student is rated as
lower in ability. Furthermore, explicit information about effort has the
predicted effects on the ability ratings. When subjects learn that the teacher
considered both children to have exerted high effort (left portion of the
figure), the difference in ability ratings is relatively small and no longer
significant . The central portion of the figure compares the control condition
with the condition in which the explicit effort information is congruent with
the effort information, otherwise implicitly contained in the feedback. The
figure shows that the ability ratings in both conditions do not differ.
However, when the explicit information on effort contradicts the effort
information implicity contained in feedback (right portion of the figure),
the ability ratings are the reverse of those in the control condition. This
is manifested in a strong interaction effect in a two-way ANOVA
(feedback x condition) with repeated measures on feedback.
In sum, these results show that inferences about ability estimates based
on feedback can be modified in a systematic fashion by explicit information
about effort. Inferences about the actor’s effort thus seem to play a crucial
role in forming conclusions about ability estimates on the basis of feedback.

RESTRICTIVE CONDITIONS
Surely praise and criticism do not always lead to conclusions about the
evaluator’s estimate of an actor’s ability; this occurs only under certain
conditions. All studies described above that led to such conclusions had two
features in common: (1) The subject learns that two children achieved
identical results; and ( 2 ) that subsequently both children received different
reactions from the teacher. Without information about identical perform-
ances, the subjects might have reached another conclusion. For example,
they could have assumed that, because of high effort, the child being praised
had attained better results than the child who received a neutral reaction.
But in the eyes of the teacher both children might have had equal ability.
268 WULF-UWE MEYER

I HE: Information of High Effort

Praise Neutral Praise Neutral Praise Neutral

Teacher’s Feedback
Figure 9.3 Mean ability ratings as a function of teacher feedback and information
about effort expenditure. From Meyer, 1984. Copyright 1984 by Verlag Hans Huber.
Reprinted by permission

Moreover, had the questionnaire contained the information that the

teacher showed identical (instead of different) reactions to both students,
e.g. had praised both for success at an easy task-or, as an extreme case,
had the subjects known that the teacher praised every child in the class for
success at an easy task-then praise would again, presumably, not have led
to conclusions about the teacher’s ability assessments. In such a case, praise
would probably lead to inferences about specific characteristics of this
teacher, e.g. that the teacher was particularly friendly or supportive, rather
than about specific characteristics of the pupils as perceived by the teacher.
Thus, the degree to which evaluative behavior provides information about
the evaluator’s estimate of the actor’s ability seems to depend on the
distinctiveness (Kelley, 1967) of the evaluative reactions. The probability of
an ability inference increases the more the evaluator’s praising, or criticizing,
of the actor, differs from the evaluator’s reactions to other individuals for
identical performances (high distinctiveness).
Another important prerequisite that we have explored in several experi-
ments is the evaluator’s knowledge of the actor’s ability. Praise and criticism
will be connected with an ability assessment only when we believe that the
evaluator has already formed a notion about the actor’s ability. Otherwise,
feedback will be attributed to other sources, such as liking or disliking of
the actor.
Meyer and Ploger (1979) varied knowledge of the actor’s ability in the
PARADOXICAL EFFECTS OF PRAISE A N D CRITICISM 269

following way. Subjects had to imagine that they were students in a

classroom. In one experimental condition, subjects were told that they had
been in the teacher’s class for a long time. Subjects in a second condition
learned that a new teacher had just come to the school who was now
teaching in the class for the first time. We assumed that the subjects would
attribute knowledge about their own ability to the familiar teacher, but not
to the unfamiliar teacher.
Subjects were then told that the teacher brought to the class several sets
of problems varying in difficulty. First, according to the cover story, the
teacher gave each child a set of 10 easy problems to solve. After the allotted
time for solving these problems had elapsed, the teacher looked at the
subject’s sheet and at the sheet of a classmate. The teacher noted that both
had correct answers on 9 out of 10 problems. Half of the subjects in each
condition (“familiar teacher” and “unfamiliar teacher”) were told that the
teacher had criticized the classmate for the outcome and praised the subject
for the identical performance. The other half in each condition were told
that the teacher had praised the classmate and criticized the subject.
Perceived ability was assessed by predictions of future success at similar
tasks. Subjects were told that the teacher would hand out four additional
sets of 10 problems classified as easy, moderately difficult, difficult, and
very difficult. Subjects indicated for each set whether they expected to solve
more (positive scores), the same (zero), or fewer problems (negative scores)
than their classmate. They were also asked to indicate how much the teacher
liked them and their classmate.
The mean expected differences are summarized in Figure 9.4. The left
portion of the figure shows that praise and criticism from a familiar teacher
have a strong influence on predictions of success. When criticized by a
familiar teacher while the classmate is praised, subjects expect to solve more
problems correctly than the classmate. But if the reverse pattern of feedback
is given, subjects expect to solve fewer problems correctly. It is interesting
to note that in both feedback conditions the expected differences in correct
solutions between oneself and the classmate do not appear in the easy task
condition, but then increase with higher task difficulty. Meyer and Ploger
attributed this finding to the fact that in easy, but not in difficult tasks, low
ability can be compensated by high effort.
To explain this in more detail, let us first consider the condition in which
the subject (oneself) was praised while the classmate was criticized. This
condition was expected to make the subject believe that the teacher estimates
the subject’s ability as low and the classmate’s ability as high. In easy tasks,
the subject can compensate his or her low ability by high effort. Therefore,
at this level of difficulty the subject expects to solve as many problems as
the classmate (the mean difference is near 0). In difficult tasks, however,
low ability cannot be compensated by high effort. Therefore, in these tasks
270 WULF-UWE MEYER

0 Onesetf: Probe. Chasmate: Criticism

o Oneself: Critkiam. CkarcMte: Raire

Familiar Teacher Unfamiliar Teacher

-- Easy Moderate Difficult Very

Difficult
Easy Moderate Difficult Very

Task Difficulty
MfficUlt

Figure 9.4 Mean differences in expected correct solutions for oneself and a classmate
as a function of feedback, familiarity with teacher, and difficulty of tasks. From
Meyer & Ploger, 1979. Copyright 1979 by Ernst Klett. Reprinted by permission

the subject expects to solve fewer problems than the classmate. Now consider
the condition in which the subject (oneself) was criticized while the classmate
was praised. This condition was supposed to make the subject believe that
the teacher estimates the subject’s ability as high and the classmate’s ability
as low. With the easy tasks, the classmate can compensate his or her low
ability by high effort. Therefore, in these tasks the subject expects no
difference between the number of own correct solutions and that of the
classmate. In the difficult tasks, however, the subject expects to solve more
problems because here the classmate cannot compensate his or her low
ability by high effort. In sum, these results show that praise and criticism
from an evaluator who apparently knows the actors’ ability can lead to
conclusions about the evaluator’s ability estimate-thereby influencing
predictions about future performance.
The right portion of Figure 9.4 shows that differential feedback from an
unfamiliar teacher does not affect performance expectations. However,
feedback in this condition influences judgments about the evaluator’s liking:
PARADOXICAL EFFECTS OF PRAISE A N D CRITICISM 27 1

if the subject is praised while the classmate is criticized, the subject believes
that the teacher likes him or her better. But if the subject is criticized and
the classmate is praised, the subject believes the fellow classmate is liked
better. On the other hand, in the condition “familiar teacher,” differential
feedback causes no significant differences in the likeability ratings.
Praise and criticism do not invariably lead to conclusions about an
evaluator’s assessment of the actor’s ability. The results of this experiment
show that one prerequisite for such conclusions is the possibility of ascribing
knowledge about the actor’s ability to the evaluator.

EFFECTS ON SELF-PERCEIVED ABILITY

After having found repeatedly that praise and criticism lead to inferences
about ability assessments by the evaluator, we went on to examine the
extent to which evaluative feedback affects the actor’s self-perception of
ability (Meyer, Engler & Mittag, 1982; Meyer, Mittag & Engler, 1986). For
this purpose we arranged laboratory experiments and informed the subjects
that they were to simulate teacher-pupil interactions, the focus being on
praise and criticism. Two of the four male participants in each experimental
session were assigned the role of “teachers”; the other two the role of
“students.” The “teachers” were experimental confederates.
In the Meyer et al. (1982) experiment, the two students initially underwent
what was said to be a test of “discrimination ability.” After the test was
completed, both teachers allegedly scored the tests in one condition to
induce the subjects to believe that the teachers knew their ability. In another
condition, the teachers did not score the tests. We expected that subsequent
feedback from the teachers would influence the students’ self-perceived
ability only when the students believed that the teachers knew their ability.
In both conditions the students then worked at an identical task which
was very similar to the alleged test. The experimenter asked the teachers
to give each student feedback after completion of the task. For that purpose
each teacher had four cards with the following labels: Praise, No Praise,
Criticism, No Criticism. The teachers were to convey their reactions by
choosing one of the cards and by showing it to the students.
Both subjects were given failure feedback for the task. Subsequently, one
of the two subjects received “criticism” from both teachers, while the other
received “no criticism”. The subjects then indicated on a scale how they
assessed their own performance on the test of ability in comparison with
that of the other student. These ratings were used as the measure of
perceived ability.
In the test-not-scored condition, there were no differences between the
ratings of those subjects who were criticized and those who were not
272 WULF-UWE MEYER

criticized. In the test-scored condition, however, quite distinct differences

appeared as a function of feedback. Subjects who were not criticized believed
that their own test performance was worse than the performance of the
other student; subjects who were criticized considered their performance
better in comparison with the other student.
Thus, evaluative behavior affected self-perceived ability in the actors only
if they could assume that the evaluators knew their ability. Unsystematic
observations during the experimental sessions also indicated that many
subjects showed rather strong (facial) emotional reactions to receiving no
criticism. They seemed to be worried and disappointed. Therefore, in a
second study, Meyer, Mittag & Engler (1986) also assessed the subjects’
affective reactions to the feedback received.
The experimental design and the procedure used in this study were very
similar to the preceding experiment (Meyer, Engler & Mittag, 1982),
although there were some differences between the two studies. The major
divergence concerned the induction of feedback. In the Meyer, Mittag &
Engler (1986) experiment, subjects received feedback not only for failure,
but also for success. One group of subjects was praised for success at an
easy task and not criticized for failure at the difficult task. A second group
received no praise for success at an easy task, but criticism for failure at
the difficult task. Thus, in each group two variants of feedback were
combined that led, in the simulational studies mentioned above, to the
belief that the evaluator estimated the actor’s ability as low (praise for
success, neutral reactions after failure) or high (neutral reactions for success,
criticism for failure). As in the Meyer, Engler & Mittag (1982) experiment,
the type of feedback factor was crossed with two levels of subjects’ belief
about whether the “teachers” had knowledge of the subjects’ ability (test-
scored condition) or had no such knowledge (test-not-scored condition).
Mean ratings of test performance are shown in the left-hand portion of
Figure 9.5. The structure of these ratings is similar to that found by Meyer,
Engler & Mittag (1982). When subjects could assume that the teachers knew
about their ability (test-scored condition), subjects who received praise for
success and no criticism for failure estimated their own test performance to
be lower, relative to the other student, while subjects who were not praised
for success, but criticized for failure estimated their own performance to be
relatively higher. (Note that criticism is labeled as blame in Figure 9.5). On
the other hand, when the teachers apparently had no knowledge of the
subjects’ ability (test-not-scored condition), differential feedback did not
affect the performance ratings.
The subjects’ affective reactions to the feedback are shown in the right-
hand portion of Figure 9.5. These ratings are based on the difference
between the scores of two positive affects (joy, confidence) and three
negative affects (anger, dejection, disappointment). When the subjects
PARADOXICAL EFFECTS OF PRAISE A N D CRITICISM 273

FEEDBACK RECEIVED

Figure 9.5 Mean performance ratings and mean affect ratings as a function of
feedback and test-scoring. From Meyer, Mittag & Engler, 1986. Copyright by
Guilford Publications, New York. Reprinted by permission

thought that the teachers were aware of their ability (test-scored condition),
then not being praised for success but criticized for failure led to a more
positive affective state than being praised for success and not being criticized
for failure. Within the test-not-scored condition, the reverse trend occurred.
However, the difference in this condition was not significant. As can be seen
from Figure 9.5, the largest difference exists between the test-scored and the
test-not-scored condition for subjects who were not praised for success but
criticized for failure: when the teachers apparently knew about the subjects’
ability, this feedback triggered predominantly positive affect in the subjects.
On the other hand, when the teachers had apparently no knowledge about
the subjects’ ability, the same feedback led to predominantly negative affect.
These results provide rather strong support for our initial contention that
praise and criticism, under certain circumstances, may have consequences
that seem paradoxical from a reinforcement perspective, which ascribes to
positive and negative verbal feedback corresponding positive and negative
affects. When the subjects could assume that the evaluators knew about
their ability, then receiving praise for success and no criticism for failure
led to the inference that one’s performance on a test of ability was low. At
274 WULF-UWE MEYER

the same time, receiving no praise for success but criticism for failure gave
rise to the belief that one’s test performance was high and also triggered
predominantly positive affect. Apparently, praise and criticism provided the
recipients with information about how the evaluators were estimating their
ability. This inferred opinion of others then influenced self-perception of
ability and affective reactions.

FEEDBACK AND OTHER ABILITY CUES

In the introductory part of this chapter, I mentioned that, in addition to
praise and criticism, other behaviors may function quite unintentionally as
high-ability or low-ability cues. These behaviors include helping, emotional
reactions, and the assignment of tasks. Our own research showed that
unsolicited help, communicating pity following failure, and the assignment
of easy tasks may function as low-ability cues, while withholding help,
communicating anger following failure, and the assignment of difficult tasks
may function as high-ability cues (for a summary see Meyer, 1984; see also
Graham, 1984; Graham & Barker, 1990; Rustemeyer, 1984; Weiner et al.,
1982). This is due to the circumstance that these behaviors-similar to praise
and criticism-are perceived to reflect particular attributions (see Weiner,
1986). Helping and pity are more likely to occur when the need or failure
of a person is attributed to uncontrollable causes such as low ability, and
easy tasks will be assigned if the ability or performance level of the recipient
is perceived to be low. On the other hand, withholding of help and anger
are more likely when the need or failure of a person is attributed to
controllable causes such as insufficient effort, and difficult tasks will be
assigned if the ability or performance level of the recipient is perceived to
be high. Thus, these behaviors may provide the recipient with information
about how others are estimating his or her ability (cf. Figure 9.1).
If feedback, helping, emotional reactions, and task assignment can indeed
function as cues for inferring low or high ability, there should then be
systematic co-occurrences among these behaviors. For example, if a teacher’s
praise for success leads to the inference that the teacher estimates the pupil’s
ability as low, then this teacher should be expected to show further reactions
to the pupil that are consistent with his or her estimate of low ability:
unsolicited help, pity following failure, and assigning easy tasks. Meyer et
al. (1988) explored this possibility in a questionnaire study using students
in their late teens and teachers as subjects.
Each participant was given one out of four short scenarios. Each scenario
described a teacher’s reaction to two students:
1. Two students succeed in solving an easy task. One student is praised,
the other student receives a neutral reaction.
PARADOXICAL E F F E n S OF PRAISE A N D CRITICISM 275

2. A teacher gives the students an intermediate-difficulty problem to solve.

While working at the problem, one student receives unsolicited help by
the teacher, and the other student is not helped.
3. A teacher assigns a very easy math problem to one of two students,
but not to the other.
4. Two students fail at a difficult task. The teacher becomes angry about
one of the students, but not about the other.

After reading the scenario, subjects indicated on a seven-point scale the

probability of the occurrence of three teacher behaviors in other school
situations. These behaviors corresponded to those described in the three
scenarios not provided to the subject. Thus, for example, subjects who read
the praise (vs. neutral reaction) scenario estimated for each of the two
students the probability that the teacher would: (a) offer unsolicited help at
an intermediate difficulty problem; (b) assign a very easy problem; and
(c) become angry about failure at a difficult task. Alternatively, subjects
who read the help (vs. no help) scenario estimated for each of the two
students the probability that the teacher would: (a) praise for success at an
easy task; (b) assign a very easy problem; and (c) become angry about
failure at a difficult task. Each scenario was rated by 14 teachers and 23
pupils.
The mean ratings are shown in Table 9.2. The student who was described
in the scenario as receiving praise, unsolicited help, a very easy task, or
anger is referred to as “Yes”; the matched pupil not receiving these reactions
is referred to as “No.” The table shows that the ratings for the two students
differ reliably (with only one exception) within each scenario. Furthermore,
teachers and students came to essentially the same judgments about the
behavior of a hypothetical teacher. Altogether, the structure of results is
highly consistent: the probability of behavior indicating an estimate of low
ability (praise, help, assignment of an easy task) is always higher for that
student whose ability would be rated as low according to the information
contained in the scenarios. At the same time, the probability of behavior
indicating an estimate of high ability (anger about failure) is always rated
as higher for that student whose ability would be rated as high according
to the information in the scenarios. For example, in the praise scenario the
praised (low ability cue) student is rated to be the target of teacher behaviors
indicating an estimate of low ability (unsolicited help, getting an easy task);
at the same time, the neutral feedback (high ability cue) is rated to be the
target of teacher behavior indicating an estimate of high ability (anger about
failure). Or, in the anger scenario, the student who received no anger (low
ability cue) is rated to be the target of behaviors indicating an estimate of
low ability (praise, help, easy task).
In sum, there exist-from the perspective of teachers and of students-
Table 9.2 Mean probability ratings of teacher behaviors (dependent variables) in four scenarios

Subjects Dependent variable Scenario

Praise Help Easy Task Anger
Yes No Yes No Yes No Yes No

Praise - - 5.57 2.21' 5.71 2.85' 2.78 5.28'

Help 4.85 2.50" - - 5.00 2.21' 2.71 5.2Ic
Teachers - -
Easy task 5.35 2.76" 5.42 2.71' 2.78 5.50b
Anger 2.50 4.2gb 1.21 2.92' 1.64 3.42 - -
Praise - - 5.43 2.56' 5.13 3.04' 3.08 5.21'
Help 4.91 2.04' - - 4.78 3.17' 2.26 4.86'
Students - -
Easy task 6.17 2.21' 5.56 2.69 2.95 4.91"
Anger 2.26 4.91' 2.04 3.69 2.04 4.43' - -
Note: Subscripts indicate the significance level between the "Yes" and "No" means for each dependent variable ( I tests, two-tailed). p < 0.05;
"p < 0.01; =p < 0.001. From Meyer, Bedau & Engler, 1988. Copyright 1988 by Verlag Hans Hubcr. Reprinted by permission.
PARADOXICAL EFFECTS OF PRAISE A N D CRITICISM 277

quite systematic co-occurrences of teacher behaviors that can function

unintentionally as cues for inferring high or low ability. Therefore, one
might speculate whether in classroom situations (and in other contexts)
there exists a syndrome of behaviors signaling to a student that the teacher
estimates his or her ability as high or low.

AGE-RELATED DIFFERENCES
All the studies reported thus far in this chapter used adults as subjects.
Interestingly, however, the ability inferences from praise and criticism which
are typical for adults are not found in younger children. In their first years
at school, children generally arrive at opposite conclusions. For instance,
Meyer el al. (1979) used questionnaires to describe two male students who
successfully solved an easy task. One student was praised by the teacher;
the other student was told only that his answer was correct. Subjects
(students aged 8-19 years) were asked whom the teacher considered to be
more able. As can be seen from Figure 9.6, most of the children aged 8-9
years considered the praised student to be more able. With increasing age,
this response was chosen less often. Finally, most of the students aged 14-16
years and 17-19 years considered the neutral-feedback student to be more
able.
Ploger (1980), Barker and Graham (1987), and Hom (1991) found similar
age-dependent differences. Ploger and Barker and Graham examined not
only success outcomes and judgments about ability, but also failure outcomes
and judgments about effort. Subjects of varying ages (Ploger: 6-17 years;
Barker and Graham: 4-12 years) were given information about two students
who received either praise vs. neutral feedback for success or criticism vs.
neutral feedback for failure. The subjects then made inferences about the
ability and effort of the two students. Ploger, as well as Barker and Graham,
found marked age-dependent differences in the judgments about ability as
a function of feedback. The youngest children inferred high ability from
praise for success and low ability from criticism for failure. The oldest
children made the opposite inferences, they considered the praised student
to have low ability and the criticized student to have high ability. However,
such age-related reversals were not found in the judgments of effort. In all
age groups, the praised student was judged higher in effort and the criticized
student was judged lower in effort than were their neutral-feedback
counterparts.
Thus, in the youngest children, the judgments of ability and effort co-
varied positively: the praised student is the one who has more ability and
at the same time has expended more effort. In the case of failure, it is the
neutral-feedback student who has more ability and at the same time has
278 WULF-UWE MEYER

I
Praised Child More Able

n "Neutral" Child More Able

8 - 9 9 - 10 10 - 14 14 - 16 17 - 19
Age of Subjects

Figure 9.6 Frequency (percentage)of ratings that the praised child or the neutral-
feedback child was considered more able by the teacher, as a function of subjects'
age. From Meyer, 1984. Copyright 1984 by Verlag Hans Huber. Reprinted by
permission. Data from Meyer e l al., 1979. Copyright 1979 by the American
Psychological Association, Inc.

given more effort. In older children, however, ability and effort have a
compensatory (negative) relationship: praise for success and neutral feedback
after failure both indicate high effort and at the same time low ability. This
is also reflected in the correlations between the effort and ability judgments,
found by Barker and Graham (1987). The coefficients for success or failure,
respectively, were: 4 to 5-year-olds, 0.84 or 0.86; 8 to 9-year-olds, 0.24 or
0.55; and 11 to 12-year-olds: -0.79 or -0.82.
Ploger (1980) and Barker and Graham (1987) explained these age-related
differences in the use of praise and criticism as cues for inferring ability in
the following way. Even the youngest subjects are able to infer the amount
of effort from information about praise and criticism. However, because
very young children perceive a positive, rather than compensatory, relation
between effort and ability (see also Kun, 1977; Nicholls, 1978), they make
their inferences of ability parallel to the extent of effort. In older subjects
this is not the case, because they understand the compensatory relation
between effort and ability. Thus, Ploger's (1980) and Barker and Graham's
(1987) interpretation assumes that the inferential processes in younger
children follow the same pattern as those of adolescents and adults (see
PARADOXICAL EFFECTS OF PRAISE A N D CRITICISM 279

Figure 9.2). An inference of effort is made on the basis of the first

attributional principle (evaluative principle); a second principle then further
permits an ability inference. This latter principle is, however, not the same
in younger children as in adolescents and adults. In younger children it
consists of a positive covariation of effort and ability; Kun (1977) labeled
this phenomenon the “halo schema.” In adolescents and adults the principle
consists of the compensatory schema.
There may well be an alternative explanation for the age-related differences
considered here: younger children might arrive at their characteristic
inferences of effort and ability on the basis of a feedback-oriented judgment
schema, in which the two attributional principles are not involved. This
schema contains the conviction that the positive or negative feedback of an
authority figure indicates a general assignment of positive or negative
properties to the actor. Thus, when the two properties, ability and effort,
are brought to the attention of the subjects via questionnaires, then positive
feedback after success would be rated as indicating that the actor is perceived
as having the positive characteristics of high effort and high ability (which
results in a positive covariation of effort and ability). The feedback-oriented
judgment schema thus leads to direct conclusions about ability from the
feedback, without involving a mediating effort inference. To be sure, Le6n-
Villagrii, Meyer and Engler (1990) reported evidence for the operation of
such a schema in 6 to 9-year-old children when inferring ability from praise.
However, because there are only very few studies examining the basis of
the well-established age-related differences in the ability inferences, the
explanatory power of these conceptions (halo schema vs. feedback-oriented
schema) is an open question.

SOME IMPLICATIONS
The evidence presented in this chapter showed that praise and other
behaviors (pity, help, assignment of easy tasks), which are frequently
considered to be positive or socially desirable, can have negative consequences
because they may quite unintentionally convey low-ability messages. On the
other hand, criticism (as well as anger following failure, withholding help,
and assigning difficult tasks), which are frequently considered to be negative
events, can imply high-ability messages. However, it is not advocated here
that, for example, it is always better to withhold than to give praise or help.
This would be absurd, in that these events do not invariably have positive
or negative consequences. With respect to praise and criticism, 1 have shown
here that it depends on the context, and on the recipient’s cognitive maturity,
whether and what kind of ability messages are conveyed by these reactions.
Although the attributional analysis of praise and criticism presented here
280 WULF-UWE MEYER

is primarily based on simulational studies and laboratory investigations,

these findings should have applications to education, welfare, and other
segments of our society. When considering the classroom, there is, first,
evidence showing that teacher praise and criticism are determined by the
teacher’s perception of a pupil’s ability. Boteram (1976), Fischer (1982),
and Weinstein (1976) observed interactions between teachers and students
in the classroom; Horn (1984) assessed coaches’ feedback following the
performances of juvenile athletes. The results of these studies were quite
uniform with respect to praise: more praise was directed at students and
athletes whom their teachers and coaches perceived to be low in ability than
at students perceived as high in ability; even though the high-ability students
performed better than the low-ability students. Furthermore, proportionally
more critical comments were directed toward students perceived to be high
in ability than those perceived to be low in ability (Weinstein, 1976). Second,
there is some evidence suggesting a causal relationship between teachers’
or coaches’ praise and criticism and students’ self-perception of ability
(Horn, 1985; Tacke & Linder, 1981). Horn (1985) found praise to be
negatively associated and criticism to be positively associated with gains in
the perceived competence of juvenile athletes:

. . . praise in this study may have been a negative coaching behavior because
it was given inappropriately and non-contingently, thereby informing certain
players that their coaches held low expectations for them. Players who received
relatively high frequencies of criticism for skill errors, however, may have
perceived such an evaluation to be an indication that their coaches attributed
their failure to lack of effort and that the coach expected them to perform at
a higher level, thus facilitating higher perceptions of competence in these
players. (p. 183)

Concerning areas beyond the classroom, one might speculate that more
reward for success (especially at easy tasks), lack of criticism for failure,
and unsolicited help, pity, and the assignment of very easy tasks is directed
at the handicapped, at older people, and at females. These behaviors may
be among the factors that mediate the low self-concepts of ability that are
frequently found in these groups.
I believe that in our everyday social interactions indirect and unintentional
ability messages are much more frequent than direct verbal communications
of how one estimates another person’s ability. Direct communications will
be suppressed, especially when the other’s ability is perceived to be low.
One will, for example, rarely say to a student or to a handicapped person
“You are unable,” because one does not want to hurt the target person’s
self-esteem. However, in spite of the intention not to cause harm, ability
estimates will be conveyed quite subtly and unknowingly and can be
important determinants of the target person’s self-perception of ability.
PARADOXICAL EFFECTS OF PRAISE A N D CRITICISM 281

ACKNOWLEDGMENTS
I wish to express my appreciation to Robert Wicklund and Margaret
Woodruff for their help in writing the manuscript in English. I also thank
Friedrich Forsterling, Daniele Kammer, Michael Niepel, Achim Schutzwohl,
Joachim Stiensmeier-Pelster, and Robert Wicklund for helpful comments
on the manuscript.

REFERENCES
Barker, G., & Graham, S. (1987). Developmental study of praise and blame as
attributional cues. Journal of Educational Psychology, 79, 62-66.
Blickle, G. (1990). Leistungssanktionen und Fahigkeitseinschatzungen: Der Ver-
stehensvorgang bei scheinbar paradoxalen Wirkungen von Lob und Tadel. Sprache
& Kognition, 9, 113-129.
Blickle, G., & Groeben, N. (1988). Gegen einen objektivistisch halbierten
Kognitivismus. Kognitiv-konstruktives Sprachverstehen und nicht-paradoxale Wir-
kungen von Lob und Tadel-Teil2. Zeifichrift far Sozialpsychologie, 19,103-1 17.
Blickle, G., & Groeben, N. (1990). Fur eine angemessene Beschreibung der
kognitiven Sprachverarbeitung: Eine Replik auf Reisenzeins technizistisch halbierte
Forschungsmethodologie. Sprache & Kognition, 9, 231-233.
Boteram, N. (1976). Pygmalions Medium. Rheinstetten: Schindele.
Brophy, J. (1981). Teacher praise: a functional analysis. Review of Educational
Research, 51, 5-32.
Fischer, C. (1982). Ursachenerklarung im Unterricht. Cologne: Bohlau Verlag.
Gage, N., & Berliner, D. (1975). Educational Psychology. Chicago: Rand McNally.
Graham, S. (1984). Communicating sympathy and anger to black and white children:
the cognitive (attributional) consequences of affective cues. Journal of Personality
and Social Psychology, 47, 40-54.
Graham, S. (1990). Communicating low ability in the classroom: bad things good
teachers sometimes do. In S. Graham & V.S. Folkes (Eds), Attribution Theory:
Applications to Achievement, Mental Health, and Interpersonal Conflict. Hillsdale,
NJ: Erlbaum, pp. 17-36.
Graham, S., & Barker, G. (1990). The down side of help: an attributional-
developmental analysis of helping behavior as a low-ability cue. Journal of
Educational Psychology, 82. 7-14.
Groeben, N., & Blickle, G. (1988). Gegen einen objektivistisch halbierten
Kognitivismus. Kognitiv-konstruktives Sprachverstehen und nicht-paradoxaleWir-
kungen von Lob und Tadel-Teil 1. Zeibchrift fur Sozialpsychologie, 19, 85-102.
Heider, F. (1958). The Psychology of Interpersonal Relations. New York: Wiley.
Hofer, M. (1985). Zu den Wirkungen von Lob und Tadel. Bildung und Erziehung,
38, 415-426.
Hofer, M.,Tacke, G., Dobrick, M., Pursian, R., Grobe, R., & Preuss, W.
(1982). Die Interpretation individualisierenden Lehrerverhaltens durch Schiiler.
Schlussfolgerungen auf selbstkonzeptbezogene Eigenschaftszuschreibungen (am
Beispiel von Tadel). Unpublished manuscript, Technical University of
Braunschweig.
Hom, H. (1991). Developmental Conceptions of Praise and Ability in Sport. Paper
presented at a meeting of the Society of Research in Child Development.
282 WULF-UWE MEYER

Horn, T.S. (1984). Expectancy effects in the interscholastic athletic setting:

methodological considerations. Journal of Sport Psychology, 6 , 60-76.
Horn, T.S. (1985). Coaches’ feedback and changes in children’s perceptions of their
physical competence. Journal of Educational Psychology, 77, 174-186.
Kanouse, D., Gumpert, P., & Canavan-Gumpert, D. (1981). The semantics of
praise. In J. Harvey, W. Ickes & R. Kidd (Eds), New Directions in Attribution
Research, Vol. 3. Hillsdale, NJ: Erlbaum, pp. 97-115.
Kelley, H.H. (1967). Attribution theory in social psychology. In D. Levine (Ed.),
Nebraska Symposium on Motivation. Lincoln: University of Nebraska Press,
pp. 192-238.
Kukla, A. (1972). Foundations of an attributional theory of performance. Psychologi-
cal Review, 79, 454-470.
Kun, A. (1977). Development of the magnitude-covariation and compensation
schemata in ability and effort attributions of performance. Child Development,
48, 862-873.
Lanzetta, J., & Hannah, T. (1969). Reinforcing behavior of “naive” trainers. Journal
of Personality and Social Psychology, 11, 245-252.
Lebn-VillagrB, J., Meyer, W.-U., & Engler, U. (1990). Fahigkeitsinferenzen
aufgrund von Lob: Eine entwicklungspsychologische Studie. Zeitschrift fur
Entwicklungspsychologie und Padagogische Psychologie, 22, 54-74.
Meyer, W.-U. (1973). Leistungsmotiv und Ursachenerklarung von Erfolg und
Misserfulg. Stuttgart: Klett.
Meyer, W.-U. (1978). Der Einfluss von Sanktionen auf Begabungsperzeptionen. In
D. Gorlitz, W.-U. Meyer & B. Weiner (Eds), Bielefelder Symposium uber
Attribution. Stuttgart: Klett-Cotta, pp. 71-87.
Meyer, W.-U. (1982). Indirect communications about perceived ability estimates.
Journal of Educational Psychology, 74, 888-897.
Meyer, W.-U. (1984). Das Konzept von der eigenen Begabung. Bern: Huber.
Meyer, W.-U., Bachmann, M., Biermann, U., Hempelmann, M., Ploger, F.-O., &
Spiller, H. (1979). The informational value of evaluative behavior: influences of
praise and blame on perceptions of ability. Journal of Educational Psychology,
71, 259-268.
Meyer, W.-U., Bedau, U., & Engler, U. (1988). Indirekte Mitteilungen uber
Fahigkeitseinschatngen in hypothetischen Lehrer-Schiiler Interaktionen. Zeit-
schrift fur Padagogische Psychologie, 2, 235-242.
Meyer, W.-U., Engler, U., & Mittag, W. (1982). Auswirkungen von Tadel auf die
Beurteilung des eigenen Leistungsstandes und auf das Interesse an Aufgaben.
Zeitschrift fur Entwicklungspsychologie und Padagogische Psychologie, 14,263-276.
Meyer, W.-U., Mittag, W., & Engler, U. (1986). Some effects of praise and blame
on perceived ability and affect. Social Cognition, 4, 293-308.
Meyer, W.-U., & Ploger, F.-0. (1979). Scheinbar paradoxe Wirkungen von Lob
und Tadel auf die wahrgenommene eigene Begabung. In S.-H. Filipp (Ed.),
Selbstkonzepr-Forschung. Problerne, Befunde, Perspektiven. Stuttgart: Klett-Cotta,
pp. 221-235.
Meyer, W.-U.. Ploger, F.-O., & Conty, M. (1980). Inferences from Praise and
Blame on Perceived Ability Estimates. Unpublished manuscript, University of
Bielefeld.
Nicholls, J. (1978). The development of the concepts of effort and ability, perception
of own attainment, and the understanding that difficult tasks require more ability.
Child Development, 49, 800-814.
OLeary, K., & O’Leary, S. (1972). Behavior modification with children. In K.
PARADOXICAL EFFECTS OF PRAISE A N D CRITICISM 283
OLeary & S. O’Leary (Eds), Classroom Management: the Successful Use of
Behavior ModiFcation. New York: Pergamon, pp. 1-48.
Ploger, F.-0. (1980). Sanktionen und Begabungswahrnehmung: Allgemeinpsycholo-
gische und entwicklungspsychologische Aspekte. Unpublished diploma thesis,
University of Bielefeld.
Reisenzein, R. (1990a). Gegen eine Fehlinterpretation des attributionstheoretischen
Modells scheinbar paradoxer Effekte von Lob und Tadel. Sprache & Kognition,
9 , 130-142.
Reisenzein, R. (1990b). Die Aussensicht, die Innensicht, und die Aussensicht der
Innensicht: Antwort auf Blickle und Groeben. Sprache & Kognition, 9, 234-231.
Rheinberg, F. (Ed.) (1988). Thernenschwerpunkt: “Paradoxe Effekte” von Lob und
Tadel. Zeitschrift fur Pudagogische Psychologie, 2,(4). Special Issue.
Rustemeyer, R. (1984). Selbsteinschatzung eigener Fahigkeit-vermittelt durch
die Emotionen anderer Personen. Zeitschrift fur Entwicklungspsychologie und
Padagogische Psychologie, 16, 149-161.
Tacke, G . , & Linder, F. (1981). Der Einfluss individualisierenden Lehrerverhaltens
auf das Selbstkonzept von Schiilern. Zeitschrift fur Entwicklungspsychologie und
Padagogische Psychologie, 13, 190-193.
Weiner, B. (1986). A n Attributional Theory of Motivation and Emotion. New York:
Springer-Verlag.
Weiner, B., Graham, S., Stern, P., & Lawson, M. (1982). Using affective cues to
infer causal thoughts. Developmental Psychology, 18, 278-286.
Weiner, B., & Kukla, A. (1970). An attributional analysis of achievement motivation.
Journal of Personality and Social Psychology, 15, 1-20.
Weinstein, R. (1976). Reading group membership in first grade: teacher behaviors
and pupil experience over time. Journal of Educational Psychology, 68, 103-1 16.