A word recognition model involving four processing stages was used, and tests of various word recognition strategies were administered to 25 fourth graders and 25 college students. The model included the following stages: (1) using information in a passage; (2) generating hypotheses from what the next word might be; (3) testing these hypotheses using cues such as partial perception of letters, word length, etc.; and (4) accepting or rejecting the hypotheses. A scientific prototypal two-channel tachistoscope and 10 word-pairs were used to test speed of word recognition of both adults and children. Then the word recognition strategies test was given. Thirty adjective-noun pairs were typed on index cards such that the first word was complete and the second word had missing letters indicated by dashes. Each card was flashed twice to all subjects. The subject's first report of the second word was recorded, as well as his response certainty. For each of five treatment conditions analysis by t-test indicated that adults were significantly faster in recognition; in the absence of a recognition, adults also reported a significantly higher proportion of correct partial perceptions at significantly faster speeds. References are included. (AW)