To investigate the forgetting of dreams a patient, skilled in recording them for analytic purposes, was investigated as follows. He was awakened at the end of every R.E.M. period and recorded his dream into a tape recorder. He went to sleep at the same time each night and was also awakened in the morning at the same time when he recalled what he could of the dreams. A forgotten dream was one that had been recorded on tape but was not recalled in the morning. Intensity ratings were devised to supplement length and number of dreams per night, and the items analysed statistically.

The results, taking .05 as significant for a two-tailed test, supported the theory that such variables as recency, intensity, amount of material and intraserial interference determine remembering. This method gave no evidence that forgetting was due to repression but the authors observe that it may have operated in both the original recorded and recalled dreams themselves.