An analysis of a multiyear General Social Survey data set supports pre
vious findings (Gaddy 1986; Gortmaker et al. 1990) that, with rigorous
controls in place, the negative correlation between television viewin
g and achievement disappears. Results of a hierarchical regression pro
cedure reveal that television viewing is not a significant predictor o
f adult vocabulary, over and above demographic variables and newspaper
readership. But the findings suggest that new methods of measuring te
levision consumption may be needed. Based on 1993 GSS data, the analys
is reveals that frequent viewing of certain types of television progra
mming is related positively and significantly to adult vocabulary scor
es. The finding highlights the need for new methods of measuring telev
ision consumption, for the traditional hours-per-day method of measuri
ng television viewership may be preventing researchers from understand
ing the true relationship that exists between television and learning.