. 2003;27(1):30–38.

Table 1.

Drinking Trends From Repeated Cross-Sectional Surveys—Examples of Measures not Available From Aggregate-Level Data

	1984 (n = 5,221)	1990 (n = 2,058)	1995 (n = 2,178)	χ²^† 1984 vs. 1990	χ² 1990 vs. 1995
All respondents, % (SE)
Current drinking	69.4 (1.6)	65.0 (1.4)	64.6 (1.6)	4.04^*	0.03
Wine	51.2 (1.8)	43.6 (1.5)	42.7 (1.9)	10.65^**	0.20
Beer	51.5 (1.3)	45.2 (1.4)	48.0 (1.6)	9.61^**	2.19
Spirits	51.8 (1.8)	43.5 (1.3)	42.6 (1.7)	13.85^***	0.07
Weekly drinking	35.9 (1.5)	29.0 (1.2)	29.2 (1.3)	13.90^***	0.12
5+ drinks ever in prior year	30.0 (1.2)	28.6 (1.2)	27.6 (1.4)	0.66	0.42
5+ drinks weekly in prior year	6.1 (0.6)	3.9 (0.5)	4.5 (0.6)	8.66^**	0.93
Current drinkers, mean (SE)				t^†† 1984 vs. 1990	t 1990 vs. 1995
Total drinking days	109.7 (4.6)	82.9 (3.9)	87.7 (3.9)	4.00^***	0.05
Wine	39.8 (2.5)	39.3 (3.0)	39.5 (3.0)	0.13	0.05
Beer	95.8 (4.1)	72.2 (3.9)	75.4 (3.6)	4.19^***	0.59
Spirits	34.1 (1.9)	31.5 (1.9)	26.2 (1.9)	0.98	1.98^*
Total heavy drinking days	19.3 (1.5)	13.2 (1.2)	13.2 (1.3)	2.71^**	0.07
Wine	1.9 (0.4)	1.5 (0.4)	1.0 (0.2)	0.63	0.99
Beer	13.9 (1.1)	9.4 (0.9)	10.5 (1.0)	2.74^**	0.91
Spirits	3.7 (0.5)	2.6 (0.5)	1.9 (0.3)	1.37	1.26

p < 0.05;

^**

p < 0.01;

^***

p < 0.001

^†

The chi-square statistic is used to test a hypotheses concerning the probability of whether a behavior or characteristic found in a sample—or in this case, the change in that behavior or characteristic found from one sample to another—is found to the same degree in the population as a whole.

^††

The t test assesses whether the means of two groups are statistically different from each other.

NOTE: This table is based on weighted data obtained from U.S. respondents participating in the 1984, 1990, and 1995 National Alcohol Surveys. The table displays percentages and means, as well as the standard error (SE).

SOURCE: Adapted from Greenfield et al. (2000).