Factors were investigated under acquisition and retention conditions which might be expected to counteract interference brought on by mixing sight and phonics methods. Experiment 1 dealt with training kindergarten children to attend to and encode letter pattern cues and the word's contextual cues. Subjects were trained individually in a series of four tasks: developing attention to letter cues, sight learning of letter-contrasted words, sight learning of function and content words, and sight learning of the words in sentence context. Results of Experiment 1 indicated that the systematic letter pattern training did not improve sight word learning and that the sentence context may not be a very effective vehicle for learning sight words. In Experiment 2, kindergarten children practiced sight words and phonics-based words in either list or sentence formats prior to testing and practice on a retention-transfer task. The factors investigated in Experiment 2 were found to be largely ineffective in improving word identification during reading. Tables and references are included. (Author/AW)