Krashen on rewards for reading

Some evidence that rewarding children for reading can
inhibit reading.
THE EFFECTS OF REWARD PROXIMITY AND CHOICE OF REWARD
ON THE READING MOTIVATION OF THIRD-GRADE STUDENTS.
Barbara Ann Marinak, 2004, dissertation, University of
Maryland
https://drum.umd.edu/dspace/handle/1903/1969
This is a remarkable study that presents evidence
against the use of rewards. Subjects were third
graders, classified as "average" readers. All
children were asked to read about 250 words of a
storybook, written one year below their reading level,
and were told that the experimenter wanted their
opinion of the book (design similar to McLoyd, 1979).
The experimenter told them a little about the
available books and the students choose what they
wanted to read.
After the reading, the experimenter told the child
that they had some free time and could stay in the
same room for a few minutes. The available activities
were reading more, doing a math game, or a jigsaw
puzzle.
The basic conditions were: children were offered a
reward of a book for doing the activity, a "token"
(e.g. friendship bracelet, Nerf ball, key chain, Pez dispenser), or no reward. (I am simplifying a bit).
Some children were given a choice of what book or
token, some not. This had no effect on the results. (I
present here the results only for groups given a
choice.)
The investigator observed their behavior for these few
minutes and noted whether the children selected
reading or a token as their first choice, how many
minutes they spent reading (some children who selected
a token at first did some reading later) and how many
words they read.
The results: Children given a reward of a book or no
reward overwhelmingly choose a book as their first
activity (13/15 of the book group, 11/15 of the no
reward group). Only two out of 15 of token-reward
group choose a book. Those in the book-reward and no
reward groups also spent much more time reading, and
read far more words.
An important point is that all the children liked to
read: At the end of the study all children were asked,
"If your best friend asked you what was the best or
most fun thing to do in this room, what would you tell
them?" All participants agreed that reading was the
"most fun" activity in the room.
Using books as a reward did no harm: Apparently, using
books sends the message that reading is a worthwhile
thing to do. But using tokens as rewards had a
profoundly negative effect. These results agree with
those of McLoyd (1979).
McLoyd, V. (1979). The effects of extrinsic rewards of differential value on high and low
intrinsic interest. Child Development, 50, 636-644.