Table 2 Pairwise statistical tests to determine (1) if trendline slopes for 2005–2009 DTCA data types were significantly different and (2) if magnitudes for 2005–2009 DTCA data types were significantly different after adjusting for effect of time
Statistic | TV vs. Print | TV vs. Radio | TV vs. Outdoor | TV vs. Internet | |
|---|---|---|---|---|---|
Trend | Δ (2nd minus 1st) | -$122,609,071 | -$146,138,808 | -$154,272,982 | -$167,279,709 |
SEdiff | $94,728,227 | $64,642,885 | $64,409,125 | $64,550,379 | |
t | −1.2943 | −2.2607 | −2.3952 | −2.5915 | |
df | 6 | 6 | 6 | 6 | |
p | 0.2431 | 0.0645 | 0.0536 | 0.0411 | |
Magnitude | Δ (2nd minus 1st) | -$1,706,354,882 | -$3,137,268,387 | -$3,177,916,521 | -$3,099,469,979 |
SEdiff | $140,278,912 | $115,175,353 | $117,948,373 | $123,036,591 | |
t | −12.164 | −27.2391 | −26.9433 | −25.1914 | |
df | 7 | 7 | 7 | 7 | |
p | <0.0001 | <0.0001 | <0.0001 | <0.0001 | |
Print vs. Radio | Print vs. Outdoor | Print vs. Internet | Radio vs. Outdoor | Radio vs. Internet | Outdoor vs. Internet |
$23,529,737 | -$31,663,911 | -$44,670,638 | -$8,134,174 | -$21,140,901 | -$13,006,727 |
$69,692,386 | $69,475,618 | $69,606,591 | $5,668,070 | $7,095,286 | $4,491,759 |
0.3376 | −0.4558 | −0.6418 | −1.4351 | −2.9796 | −2.8957 |
6 | 6 | 6 | 6 | 6 | 6 |
0.7471 | 0.6646 | 0.5448 | 0.2013 | 0.0247 | 0.0275 |
$1,430,913,505 | -$1,471,561,639 | -$1,393,115,097 | -$40,648,133 | $37,798,408 | $78,446,542 |
$92,111,455 | $92,526,092 | $94,212,421 | $8,601,111 | $14,628,702 | $9,106,204 |
15.5346 | −15.9043 | −14.787 | −4.7259 | 2.5839 | 8.6146 |
7 | 7 | 7 | 7 | 7 | 7 |
<0.0001 | <0.0001 | <0.0001 | 0.0021 | 0.0363 | <0.0001 |