Two experiments will be presented which assessed free recall of spoken words in Swedish (native tongue) and in English heard under different signal-to-noise (SN) ratios (+3 and +12 dB), and different reverberation time (RT, .3 and 1.2 sec). All participants encountered all eight experimental conditions (Language*SN*RT). The first experiment was run with college student (N=48) and they were run individually. In the second experiment children in grade 4 (10-11 years, N=72) took part and they were run in the regular classrooms.

Twelve wordlists in Swedish and twelve wordlists in English generated. The words were chosen according to their ranks in category norms for the two languages, and no category was the same for the two languages. The number of words in each list was 12 for the college group and 8 for grade four. The 2 x 12 wordlists were presented in counter balanced presentation orders in three blocks. Within each block order of S/N and RT was also counterbalanced. After each wordlist the participants wrote down the words they could recall. Pre-experimental measures of working memory capacity were also taken.

The basic hypotheses for the recall of the words were that working memory would be overloaded when the SN-ratio was low and the RT was long, and that SN and RT would interact with each other, with Language and with Age-group.

To compare primacy and recency effects the word lists were divided into three parts (p3rd). After each wordlist the participants typed in or wrote down the words they could recall.

The basic hypotheses for the recall of the words were that working memory would be overloaded when the SNR was low and the RT was long, and that SNR and RT would interact with each other, with Language and with Study (Grade4/College). The analyses suggest that for both groups there were expected effects of language and of SNR, but the effect of RT was smaller and only showed up in interactions.