Did Augmenting the VERB^™ Campaign Advertising in Select Communities Have an Effect on Awareness, Attitudes, and Physical Activity?

Article Outline

• Abstract
• Introduction
• Methods
  • Selecting the High-Dose Communities
  • Selecting the Comparison Sample
  • Description of the Augmented Dose 2002–2003
  • Description of the Augmented Dose 2003–2004
  • Study Design and Participants
  • Analytic Approach
• Results
• Discussion
• Acknowledgment
• Supplementary data
• References
• Copyright

Background

Although VERB^™ was designed as a national media campaign, funding and donated media time enabled more-intensive advertising and marketing in certain communities. To investigate the effect of increased advertising on physical activity outcomes, six “high-dose” communities were selected to receive more hours of advertising and additional promotional activities.

Design

Longitudinal quasi-experimental design comparing outcomes in six communities that received additional VERB marketing activities with outcomes in a comparison group that received only the national dose of advertising.

Setting/participants

Two cohorts of dyads of youth aged 9–13 years (tweens) and one parent at baseline (2002), followed for 2 years.

Intervention

During the first year of the VERB campaign, each of the six high-dose communities received 50% more advertising and conducted special campaign activities. During the second year, only four of the six communities received the larger dose of advertising and additional promotional activities because of reduced funding.

Main outcome measures

Awareness and understanding of VERB messages; attitudes about physical activity (self-efficacy, social influences, and outcome expectations); and physical activity behaviors.

Results

After 1 year, tweens in the high-dose communities reported higher levels of awareness and understanding of VERB and scored higher on the social influences scale than did tweens in a comparison group in areas that received only the national dose of advertising. After 2 years, tweens in the high-dose communities reported higher awareness and understanding of VERB, greater self-efficacy, more sessions of free-time physical activity per week, and were more active on the day before being surveyed than tweens in the comparison group who received the average national dose.

Conclusions

Providing communities with a higher dose of marketing activities and sustaining those activities over time yields more positive outcomes.

Back to Article Outline

Introduction 

From the outset, the VERB^™ campaign was designed as a national campaign that would reach U.S. youth aged 9–13 years (tweens) with messages encouraging them to be physically active every day. However, campaign planners were interested in whether conducting a higher level of campaign activities in selected communities would increase VERB's effects on the target audience. Therefore, six communities (termed high-dose communities) were selected in which to test this hypothesis. For the high-dose communities, special promotional activities were conducted and additional advertising purchased in each community. The analyses presented herein are secondary to the main VERB outcome evaluation and present a supplemental research question.

Other studies found that awareness, attitude, and behavioral outcomes vary by amount of advertising. For example, Niederdeppe¹ and Southwell², ³ demonstrated a positive association between advertising volume and awareness of the advertisement among young people targeted by national media campaigns. Snyder et al.⁴ showed that alcohol advertising expenditures per capita were positively associated with alcohol consumption by young people aged 15–26. For the truth^™ campaign (designed to discourage tobacco use among young people aged 12–17), the amount of advertising varied across media markets nationwide. Evaluation results show that high exposure to truth messages was associated with negative attitudes toward the tobacco industry and a decline in smoking prevalence.⁵, ⁶ Emery and colleagues⁷ found a similar relationship for young people exposed to state-sponsored anti-tobacco advertising. Their results show that more television advertising led to increased awareness of the anti-tobacco message, more favorable anti-smoking attitudes, and a lower likelihood of smoking.

Studies also show that activities in conjunction with television advertising lead to more positive outcomes than the outcomes resulting from television advertising alone. Two studies demonstrate that young people exposed to an in-school anti-drug curriculum in addition to an anti-drug media campaign had lower self-reported alcohol and marijuana use than had those exposed only to the advertising and not to the curriculum.⁸, ⁹ In a study to reduce cigarette smoking among adults in England, McVey and Stapleton¹⁰ systematically varied the amount of television advertising and the amount of local anti-smoking activities (e.g., public relations or grassroots activities) in four communities. They found no differences between the treatment communities (those exposed to anti-smoking activities) and comparison communities after the first year of the campaign; however, after 18 months they did find a reduction in smoking in the treatment communities compared with the comparison communities but no evidence of local activities having an additive effect over advertising alone. The researchers concluded that the effects of additional activities take a long(er) time to achieve. This conclusion is consistent with findings from Snyder and Hamilton's meta-analysis,¹¹ which found that for campaigns that did not have an enforcement message (e.g., enforcing behaviors through laws or rules), effect sizes were greater when a campaign lasted 6–18 months than when it was shorter than 6 months or longer than 2 years.

To understand better how different levels of VERB advertising would affect tweens, VERB designed a parallel study to the national outcome study.¹², ¹³ The study compared outcomes for 2 years in high-dose communities with outcomes in a comparison sample that received the national dose of advertising. We hypothesized that we would detect higher awareness of VERB, more positive attitudes toward physical activity, and more positive behavior change among tweens in the communities that received the increased advertising and promotional activities.

Back to Article Outline

Methods 

To test this hypothesis, a longitudinal quasi-experimental design was used to compare outcomes reported by tweens living in communities that received additional advertising and marketing activities with a comparison group of tweens in communities that received the national dose of advertising. This design did not allow for a separation of the additive effect of the advertising from the other marketing activities.

Selecting the High-Dose Communities 

Selecting six high-dose communities began by sorting all metropolitan statistical areas and primary metropolitan statistical areas in the U.S. by their relative population sizes and grouping them into six strata of approximately equal total population. Thus, the first stratum contained a few large communities and the sixth contained many, much smaller communities. After sorting by size, communities in each stratum were reviewed as candidates for the supplemental marketing activities with the goal that the six high-dose communities together would provide a broad demographic and geographic mix. Another factor considered was the size and cost of the media market. For example, the first stratum (largest population cities) comprised only Los Angeles, Chicago, and New York. Los Angeles was selected because it offered racial and ethnic diversity (especially Hispanic or Latino and Asian American); it had a large media market that was less expensive than New York; and it could represent the west coast geographically. Green Bay WI was selected from the sixth stratum (representing the group of smallest communities) because it had a large media market that captured a relatively high proportion of American Indians and is a northern state. The six communities selected, ranked by size, were Los Angeles, Houston, Miami, Columbus OH, Greenville SC, and Green Bay WI. VERB planners acknowledged that this group of communities did not perfectly balance all the criteria; for example, there are more southern cities than northern cities. Southern cities were preferable because a warmer climate would facilitate tweens' participation in outdoor activities that were planned for the promotional events in the first year.

After the first year of the campaign, a decrease in VERB funding required that the number of high-dose communities be reduced from six to four. Only four communities (Los Angeles, Miami, Columbus, and Greenville) continued to receive the higher dose of advertising.

Selecting the Comparison Sample 

The nationally representative sample drawn to assess the effects of VERB on the nation's tweens was not appropriate for comparison with the high-dose communities (Table 1, Table 2). Metropolitan areas were overrepresented in the high-dose communities compared with the nation as a whole, and rural areas were underrepresented. The high-dose communities had a lower percentage of white and African-American tweens and a higher percentage of Hispanic tweens than the nation as a whole. Because of the differences between the high-dose communities and the nationally representative sample, a comparison group was created from the national sample for this analysis. The comparison group consisted of an aggregate of all metropolitan areas excluding those from which the high-dose communities had been chosen and nearby counties where the target audience could have been exposed to the augmented advertising. By constructing this comparison group, the differences noted above were ameliorated. Both the high-dose communities sample and the comparison sample were then weighted to the population totals of the high-dose communities as a whole, permitting comparison of the national dose and high-dose of advertising on equivalent populations. In 2003, a comparison group was constructed for the six high-dose communities. In 2004, four of the original communities received the augmented dose of advertising; the other two communities were excluded from the analysis and the comparison group.

Table 1. Youth Media Campaign Longitudinal Survey populations, by selected characteristics, 2003

Characteristic	National population		Comparison population		High-dose population		Difference between comparison and high-dose populations	CI
	Number (thousands)	Percent	Number (thousands)	Percent	Number (thousands)	Percent
Total	20,818		16,744		16,744		0.0	0.0,0.0
Gender
Male	10,636	51	8,293	50	8,556	51	1.6	−0.5,3.6
Female	10,182	49	8,451	50	8,188	49	−1.6	−3.6,0.5
Age (years)a
10–11	8,601	41	6,749	40	6,957	42	1.2	−0.8,3.3
12–14	12,217	59	9,995	60	9,787	58	−1.2	−3.3,0.8
Race/ethnicity
White	12,921	62	9,377	56	9,453	56	0.5	−1.6,2.6
African American	3,244	16	2,221	13	2,060	12	−1.0	−2.5,0.6
Hispanic	3,540	17	4,152	25	4,299	26	0.9	−0.4,2.2
Other	1,113	5	994	6	932	6	−0.4	−2.3,1.5
Parent's education
High school or less	8,035	39	6,787	41	6,510	39	−1.7	−6.2,2.8
Some college	6,850	33	5,203	31	5,339	32	0.8	−3.1,4.7
College degree or higher	5,933	28	4,737	28	4,888	29	0.9	−2.3,4.1
Household income ($)
≤25,000	4,492	22	3,671	22	3,917	23	1.5	−1.5,4.5
25,001–50,000	5,898	28	4,446	27	4,644	28	1.2	−2.5,4.9
50,001–75,000	4,669	22	3,758	22	3,560	21	−1.2	−4.9,2.5
>75,000	5,759	28	4,870	29	4,623	28	−1.5	−4.8,1.9
Urbanicity
Urban	3,034	15	3,434	21	3,483	21	0.3	−3.0,3.6
Metro suburban	5,306	25	5,676	34	5,683	34	0.0	−3.7,3.8
Second city	3,559	17	3,344	20	2,976	18	−2.2	−5.4,1.0
Town/exurban	4,796	23	2,963	18	3,316	20	2.1	−0.5,4.7
Rural	4,122	20	1,327	8	1,286	8	−0.2	−1.9,1.4
Census region
Northeast	3,801	18	4,082	24	NA	NA	NA	NA
Midwest	4,783	23	4,117	25	5,705	34	9.5	6.7,12.3⁎
South	7,125	34	4,808	29	8,265	49	20.6	18.0,23.3⁎
West	5,109	25	3,737	22	2,774	17	−5.8	−8.3,−3.2⁎
Stratum
Nonmetro	2,713	13	NA	NA	NA	NA	NA	NA
1	2,784	13	2,774	17	2,774	17	0.0	0.0,0.0
2	2,669	13	2,870	17	2,870	17	0.0	0.0,0.0
3	2,810	14	2,739	16	2,739	16	0.0	−0.1,0.1
4	2,539	12	2,790	17	2,790	17	0.0	0.0,0.0
5	3,349	16	2,656	16	2,656	16	0.0	0.0,0.0
6	3,954	19	2,915	17	2,915	17	0.0	0.0,0.0

Note: Numbers may not add to totals because of rounding.

NA, not applicable

Table 2. Youth Media Campaign Longitudinal Survey populations, by selected characteristics, 2004

Characteristic	National population		Comparison population		High-dose population		Difference between comparison and high-dose populations	CI
	Number (thousands)	Percent	Number (thousands)	Percent	Number (thousands)	Percent
Total	62,098		10,924		10,924		0.0	0.0,0.0
Gender
Male	31,865	51	5,554	51	5,554	51	0.0	0.0,0.0
Female	30,233	49	5,371	49	5,371	49	0.0	0.0,0.0
Age (years)a
10–11	25,711	41	4,554	42	4,554	42	0.0	0.0,0.0
12–14	36,387	59	6,371	58	6,371	58	0.0	0.0,0.0
Race/ethnicity
White	38,855	62	5,306	49	5,323	49	0.2	−2.1,2.4
African American	9,652	16	1,734	16	1,734	16	0.0	0.0,0.0
Hispanic	10,530	17	3,243	30	3,243	30	0.0	0.0,0.0
Other	3,061	5	642	6	625	6	−0.2	−2.4,2.1
Parent's education
High school or less	24,024	39	4,180	38	4,335	40	1.4	−4.1,7.0
Some college	20,510	33	3,432	31	3,438	31	0.0	−4.7,4.8
College degree or higher	17,563	28	3,313	30	3,152	29	−1.5	−5.7,2.8
Household income ($)
≤25,000	13,331	21	2,437	22	2,776	25	3.1	−1.6,7.8
25,001–50,000	17,633	28	2,907	27	3,229	30	2.9	−2.4,8.3
50,001–75,000	13,882	22	2,640	24	2,259	21	−3.5	−9.3,2.4
>75,000	17,251	28	2,940	27	2,661	24	−2.6	−7.0,1.9
Urbanicity
Urban	9,043	15	3,147	29	3,147	29	0.0	0.0,0.0
Metro Suburban	15,825	25	1,240	11	1,240	11	0.0	0.0,0.0
Second city	10,637	17	3,543	32	3,543	32	0.0	0.0,0.0
Town/exurban	14,258	23	2,180	20	2,180	20	0.0	0.0,0.0
Rural	12,335	20	815	7	815	8	0.0	0.0,0.0
Census region
Northeast	11,342	18	3,431	31	NA	NA	NA	NA
Midwest	14,268	23	2,294	21	2,781	25	4.5	1.3,7.6⁎
South	21,228	34	2,527	23	5,396	49	26.3	23.3,29.2⁎
West	15,260	25	2,672	25	2748	25	0.7	−2.8,4.2
Stratum
1	7,682	13	2,748	25	2,748	25	0.0	0.0,0.0
3	7,639	13	2,739	25	2,739	25	0.0	0.0,0.0
4	8,629	14	2,781	26	2,781	26	0.0	0.0,0.0
5	7,833	13	2,656	24	2,656	24	0.0	0.0,0.0

Note: Numbers may not add to totals because of rounding.

NA, not applicable

Description of the Augmented Dose 2002–2003 

During the first year of VERB, the augmented dose in the high-dose communities occurred from October 2002 through June 2003 and had three components: (1) additional paid advertising, (2) events and promotional activities, and (3) partnerships with local groups and organizations. The additional paid advertising in high-dose communities for the first year of the campaign is detailed in Table 3. The general-market advertising (i.e., advertising targeting all U.S. tweens) purchased for television and radio during the first campaign year in each of the high-dose communities was 50% more than what would have been purchased had the community not been a high-dose community. For example, a VERB advertisement that aired on Nickelodeon 10 times per week across the nation would have aired 15 times in the high-dose communities. Out-of-home advertising (e.g., billboards, mall kiosks) was displayed in the high-dose communities. In addition to the additional general-market advertising, additional advertising was purchased in the high-dose communities where large numbers of an ethnic group resided (e.g., Hispanics, African Americans). For example, in Houston, Los Angeles, and Miami, advertisements for tweens were developed in Spanish and aired on Spanish-language television and radio from October 2002 through April 2003.

Table 3. Additional paid advertising in high-dose communities: Year 1 (2002–2003)

Plan element	Television	Radio	In school
General-market advertising
GRPs or reach	100–200 GRPs per week (national was 119 per week) Nickelodeon, Warner Bros., Disney, Cartoon Network	50 GRPs per week (National was 59 GRPs per week)	NA
Time frame	October 2002–June 2003 (37-week presence) After-school, evenings, and Saturday mornings	October 2002–June 2003 (37-week presence) Afternoon drive, evenings, and weekends	NA
Advertising targeted to Hispanics
GRPs or reach	Average of 25–35 GRPs per week in Houston, Los Angeles, and Miami (national was 10–20 per week) Univision, Telefutura, and Galavision	25–35 GRPs per week Houston: KPTY-FM KQQK-FM Los Angeles: KBUA-FM KSSE-FM KSSC-FM KSSD-FM Miami: WRTO-FM WXDJ-FM	Middle schools with ≥50% Hispanic population Book covers distributed Houston:70,800 Los Angeles: 483,000 Miami: 65,000 Custom published pieces Houston: 71,190 Los Angeles: 84,240 Miami: 10,380
Time frame	October 14, 2002–April 27, 2003 28 consecutive weeks 3pm–9pm throughout the week	October 2002–April 2003 28 consecutive weeks 3pm–9pm Fri–Tues	Once in August 2002; book covers distributed Once in March 2003; custom published pieces; distributed in ESL classes
Advertising targeted to African Americans
GRP or reach	50 GRPs per week: Miami, Houston, Los Angeles, Greenville, Columbus Stations varied by market; examples include Nickelodeon, Cartoon Network, Black Entertainment Television, and MTV	NA	Columbus, Houston, Los Angeles, Miami, Greenville: 309,102 tweens reached
Time frame	October 7–November 17, 2002 March 18–April 21, 2003 10-week schedule		Book covers distributed October 7–December 30, 2002 Custom Weekly Readers distributed March 3–April 7, 2003

ESL, English as a second language; GRP, gross ratings points.

The second component of the augmented dose comprised experiential marketing activities that included street-team marketing and participation in community events in each of the high-dose communities. Street-team marketing used youth aged 16 to 21 years to interact with tweens at high-traffic venues such as malls and theaters, talking about being physically active and distributing VERB-branded items such as temporary tattoos and wristbands.¹⁴ Street teams were used in high-dose communities by the general-market advertising agency and African-American and Hispanic advertising agencies. The general-market team interacted with tweens at 114 sites across the high-dose communities, including concerts, amusement parks, and recreation and community events. The African-American street-team vendor carried out 110 planned street-team events between August and October 2002 in the markets of Columbus, Greenville, Houston, Los Angeles, and Miami. Locations included state and county fairs, youth sporting practices, parks, malls, and recreation and community centers such as the Boys & Girls Club and the YMCA. Between August 2002 and March 2003, African-American street teams interacted with tweens at events such as Roots Heritage Festival, Miami Carnival, Healthy Black Families on the Move, and BP Houston Children's Festival.

The Hispanic street teams participated with tweens at 52 appearances in Houston, Los Angeles, and Miami. Street-team locations included malls, college and professional sporting events, parks, and markets. Events included La Feria de la Mujer, Hispanic Heritage Festival, Feria de las Americas, Parada de los Reyes Magos, Fiesta de Independencia, People en Espanol Lo Ultimo Tour, El Monte Family Festival, Latino Book and Family Festival, Noche del Grito, Fiestas Patrias Parade, church festivals, and Festival de Dia de los Muertos.

Participation in events in the high-dose communities took several forms. VERB Action Day in Los Angeles was a customized VERB event that drew over 6000 tweens (Table 4). In addition, a Nickelodeon-sponsored event was staged in each community. The Nickelodeon event, called the Wild & Crazy Kids (WACK) show was a 1-hour, live, staged version of the Wild & Crazy Kids series on Nickelodeon television. The VERB pre-event to the WACK show featured tweens participating in active games and a chance to learn about opportunities to be physically active in their communities. Across the six high-dose communities, approximately 8700 tweens attended a WACK show. VERB booths were present at festivals and fairs where attendees could play in a VERB Activity Zone and receive VERB-branded items such as Frisbees^®.

Table 4. Summary of events in high-dose communities: Year 1 (2002–2003)

WACK, Wild & Crazy Kids

The third component of the high-dose strategy was to establish a partnership with a coalition of local groups and organizations in each city. VERB staff members worked with community members in the six high-dose communities to establish a foundation of support and sustainability for VERB and physical activity that would continue after the VERB advertising campaign ended. Some community groups had coalitions already working on youth physical activity issues; other groups formed to facilitate VERB activities in their communities. The community groups generated VERB brand awareness through local media stories and events and provided assistance at the infrastructural level by addressing barriers to physical activity. For example, local chapters of Boys & Girls Clubs worked to include more physical activity in their programs. The community groups directed much of their effort in the first year to encouraging community-wide participation in the Nickelodeon-sponsored WACK event.

Description of the Augmented Dose 2003–2004 

In the second year, the components of the high-dose community strategy comprised the same three elements. However, in the four remaining communities the amount of television and radio advertising was reduced to 40% more than the community would have received if it had received only the national dose (Table 5). Advertising targeted to African Americans and Hispanics was concentrated in high-dose communities. For example, three TV advertisements developed for African-American tweens specifically featured African-American tweens and were produced in Miami, Houston, and Los Angeles. The Hispanic-targeted advertising reached Houston, Miami, and Los Angeles tweens through in-school activities and out-of-home advertising. Experiential and event marketing activities continued as well (Table 6). A Nickelodeon-sponsored tour Let's Just Play featured a VERB booth in Houston and Miami. Asian tweens were reached at the Harvest Moon Festival in Los Angeles where VERB premiums (e.g., stickers, tattoos, foot bags) were distributed to approximately 15,000 tweens.

Table 5. Additional paid advertising in high-dose communities: Year 2 (2003–2004)

GRP, gross ratings points

Table 6. Summary of events in high-dose communities: Year 2 (2003–2004)

During the second year, the campaign began to explore partnerships beyond the coalitions in the high-dose communities and one-time event participation by connecting national organizations such as Boys & Girls Clubs and YMCA to the VERB brand. The partnership team provided VERB materials to these national partners to help them reframe physical activity as an organizational or community priority, and referring partners to funding and training resources that enabled them to increase local opportunities for tweens to be physically active.¹⁵

While the augmented dose of paid media (TV, radio, and out of home) was fairly equal across the high-dose communities, more partnership efforts occurred in some communities than in others, especially in the second year. For example, the campaign hosted conferences for approximately 47 community partners in Los Angeles and 85 community partners in Houston during the second year. Both meetings recognized the accomplishments of partners who had worked with the campaign on past events, renewed the interest of existing partners, and inspired new partners to collaborate with VERB during the third year of the campaign.

Study Design and Participants 

Parent–tween dyads were randomly selected from each of the six high-dose communities and surveyed by telephone using the Youth Media Campaign Longitudinal Survey (YMCLS). Surveys were conducted from mid-April to mid-June each year. The high-dose community and national samples were combined for data collection and interviewed using the same procedures; procedures and response rates are reported elsewhere.¹² A flag on each case identified it as belonging to the national or community sample so the two samples could be separated for analysis. In 2002 for the baseline survey, youth aged 9–13 and one parent were interviewed. The dyads were followed each year for 2 years (six-community sample at baseline, n = 3217; Year 1, six communities, n = 2,771; Year 2, four communities, n = 1344; comparison sample for Year 1, n = 2088; Year 2, n = 1069).

Interviews with tweens focused on participation in and attitudes toward physical activity. Participation in specific activities during the 7 days prior to the interview was measured. Also measured was whether the activities were organized, engaged in as part of a team or structured lessons, or engaged in during free time (assessed on the basis of scales previously used in youth physical activity research).¹⁶ Scales were created to measure tweens' expectations about the outcomes of physical activity and the social influences related to physical activity. Items added to the YMCLS in 2004 allowed for the creation of a scale measuring self-efficacy with regard to being physically active. Awareness was measured by tweens' unaided or aided reports of VERB as an advertisement for kids getting active. Tweens who were aware of VERB were also asked to explain VERB briefly to confirm understanding of campaign messages. Both measures of awareness were used in this analysis. Questions to measure awareness of VERB were asked after outcome data had been collected in both 2003 and 2004. Demographic information about the tween, family, and household was gathered in the interview with parents.

Analytic Approach 

To compare outcomes for the high-dose versus the national dose, weights were developed that matched tweens in the six high-dose communities and the national comparison sample in 2003 and tweens in the four high-dose communities and the national comparison sample in 2004. Thus, the unit of analysis was all tweens in the high-dose communities group compared with all tweens in the national comparison group. Propensity scoring was used to model the systematic differences between the high-dose communities and the comparison samples. The methods of propensity scoring in the VERB outcome evaluation are explained more fully in another article in this supplement.¹⁷ The model predicted the probability of belonging to the comparison sample versus the six-community sample in 2003 and the probability of belonging to comparison sample versus the four-community sample in 2004. Based on this probability, both samples were sorted and divided into five strata to create the matching weights used in the analysis for each year.¹⁸ The matching weights were developed using a ratio adjustment where the numerator was the sum of the community weights and the denominator was the sum of the comparison sample weights. After weighting, no significant differences between the high-dose and comparison groups were evident on the important tween, parent, and household characteristics used in the analysis in either year (Table 1, Table 2).

The high-dose and comparison populations were compared on each of the outcome measures after 1 year of the campaign (except for the self-efficacy scale, because items comprising that scale were not administered until 2004), and after 2 years of the campaign. The analysis for the second year did not include participants from two of the original high-dose communities (Houston and Greenville) because the additional advertising had not continued in those communities past the first year. Therefore, only the four communities that received the augmented advertising in the second year were matched to a national comparison group for 2004. Significance was determined by a t-test of the confidence interval of the difference between the estimates for the comparison and high-dose populations.

Back to Article Outline

Results 

After 1 year, differences between tweens in the six high-dose communities and those in the comparison group were evident for three of the seven outcomes (Table 7). Awareness of the VERB campaign was higher for tweens in the high-dose population, as was awareness plus understanding of VERB messages. Tweens in high-dose communities also scored higher on the social influence scale (i.e., reported their friends and peers had positive opinions about physical activity and were likely to engage in physical activity) than did those in the comparison population. There were no differences on any of the physical activity measures. Likewise, no pattern of differences was found for racial and ethnic groups across the outcomes.

Table 7. Differences in outcome measures: high-dose communities and national-dose comparison group, 2003

Outcome measures	High-dose communities	National comparison group	Absolute difference	95% CI	Effect size (d)
Percent aware of VERB	83.7	76.1	7.6	4.9,10.3⁎	NA
Percent aware of VERB and understand VERB's messages	76.8	68.2	8.6	5.5,11.7⁎	NA
Mean score on Outcome Expectations scalea	10.06	10.02	0.03	−0.04,0.11	0.03
Mean score on Social Influence scalea	10.06	9.98	0.08	0.02,0.14⁎	0.08
Percent physically active yesterday	60.6	58.7	1.9	−1.7,5.4	0.03
Median sessions of free-time physical activityb	4.38	4.08	0.30	−0.07,0.67	0.01
Percent who participated in organized physical activityb	35.9	36.7	−0.80	−3.9,2.2	−0.02

Note: Youth were aged 10–14 in 2003. Means and medians were calculated to two decimal places; percentages were calculated to one decimal place.

In the second year, differences between the four high-dose communities that received 2 years of additional dose of advertising and the comparison group were evident for five of eight outcomes (Table 8). Awareness of VERB and awareness and understanding of VERB messages were again higher among the tweens in the high-dose communities than among tweens in the comparison population. Tweens in the high-dose communities scored higher on the self-efficacy scale, first measured in 2004, than did those in the comparison-group. No differences were found for the social influence scale.

Table 8. Difference in outcome measures for targeted tweens: high-dose communities and national-dose comparison group, 2004

Outcome measures	High-dose communities	National comparison group	Absolute difference	95% CI	Effect size (d)
Percent aware of VERB	88.3	84.6	3.7	0.6,6.7⁎	NA
Percent aware of VERB and understand messages	85.8	81.6	4.2	0.7,7.7⁎	NA
Mean score on Outcome Expectations scalea	10.06	9.98	0.08	0.00,0.16	0.08
Mean score on Social Influence scalea	9.98	9.94	0.04	−0.06,0.15	0.04
Mean score on Self-efficacy scalea	10.06	9.94	0.11	0.01,0.21⁎	0.11
Percent physically active yesterday	61.0	55.3	5.7	0.8,10.6⁎	0.12
Median sessions of free-time physical activityb	4.06	3.48	0.58	0.00,1.15⁎	0.13
Percent who participated in organized physical activityb	37.1	37.5	−0.4	−5.4,4.5	−0.01

Note: Youth were aged 11–15 in 2004. Means and medians are calculated to two decimal places; percentages are calculated to one decimal place.

In 2004, a greater percentage of tweens in the high-dose communities than tweens in the comparison group reported being physically active on the day before the interview, a relative difference of 10%. Tweens in the high-dose communities also engaged in more free-time physical activity sessions than did tweens in the comparison group.

Back to Article Outline

Discussion 

The study reported here shows that increasing the volume of VERB advertising beyond the national-level dose produced more awareness and physical activity behaviors. These findings are consistent with previous research.¹, ², ³, ⁷ Those studies also showed that additional advertising and marketing activities beyond a dose that yields intended behavioral outcomes can produce a detectable increase in awareness and an effect on attitudes and behavior. In addition, our findings show that the higher levels of awareness and understanding of VERB were achieved in both the first and second years of the campaign Attitudinal effects were found in Year 2, and similar to findings from research by McVey and Stapleton,¹⁰ behavioral effects of the VERB campaign were also achieved through sustained advertising and promotional activities. Although no behavioral effects were found after the first year of advertising, after the second year tweens in the high-dose communities reported higher levels of physical activity than did those in the national comparison group. Like McVey and Stapleton, the design did not allow for testing for an additive effect due to promotional activities. The higher levels of physical activity were important; for instance, 10% more tweens in the high-dose communities were active on the previous day than tweens in the comparison group.

Although this research design did not allow for a separation out of the relative contribution of different advertising and marketing activities, we believe that the findings may be due to the synergistic effect of the television, experiential marketing, and partnership activities. The VERB campaign messages were designed specifically to first generate awareness and interest about the meaning of VERB itself and then to entice the audience to try being physically active. We suspect that the additional activities in the high-dose communities during the first year created a stronger memory of the VERB message, priming tweens to be more receptive to subsequent campaign messages. Although additional advertising may yield additional effects, the optimal level of advertising is unknown. Moreover, the costs associated with the additional advertising will vary based on the initial advertising expenditure, amount of in-kind contributions from media partners, and costs for advertising in specific markets.

After a systematic review of physical activity interventions done for the Guide to Community Preventive Services (Community Guide),¹⁹ reviewers concluded that community-wide campaigns to increase physical activity can increase people's physical activity levels, and the Community Guide recommends such campaigns as an approach. From a public health perspective, media campaigns should work in combination with community-centered activities to have broad effects such as catalyzing coalitions that can bring about environmental or school policy changes, which result in more long-lasting effects.²⁰ Although we hypothesized that there would be synergistic and positive effects from combining additional media advertising with community-centered activities, the sample size was insufficient to allow us to discern whether VERB-related community efforts or tweens' exposure to VERB-related community events had an additional effect.

There are limitations to this study. Measurement of outcomes was based on self-report, with inherent threats to validity and potential for unreliability, especially when respondents are tweens. However, the YMCLS measures have been shown to have acceptable validity and reliability.²¹ Also, our method of selecting communities for the additional advertising resulted in the high-dose sample having different demographic characteristics than the nation as a whole. Although the results could potentially be ascribed to the communities' characteristics rather than solely to more advertising, the creation of the comparison sample and the weighting procedures to equalize the two samples mitigated this threat.

These findings suggest that an augmented advertising approach sustained over time can be effective in increasing awareness and physical activity among tweens. Future initiatives that want measurable results in a short time (e.g., 1 year) will require additional investment in augmented activities. Using these results, national health marketing campaigns on physical activity could target specific communities with greater intervention activities to yield more positive results. Such a strategy could be used to address communities where physical activity levels are low.

Back to Article Outline

The findings and conclusions in this paper are those of the authors and do not necessarily represent the views of the CDC.

No financial disclosures were reported by the authors of this paper.

Back to Article Outline

Supplementary data 

Back to Article Outline

References 

1. Niederdeppe J, Farrelly MC, Haviland ML. Confirming “truth”: more evidence of a successful tobacco countermarketing campaign in Florida. Am J Public Health. 2004;94:255–257
   • View In Article
   • MEDLINE
   • CrossRef
2. Southwell BG, Barmada CH, Hornik RC, Maklan DM. Can we measure encoded exposure? (Validation evidence from a national campaign). J Health Commun. 2002;7:445–453
   • View In Article
   • MEDLINE
   • CrossRef
3. Southwell BG. Between messages and people: a multilevel model of memory for television content. Communic Res. 2005;32:112–140
   • View In Article
4. Snyder LB, Milici FF, Slater M, Sun H, Strizhakova Y. Effects of alcohol advertising exposure on drinking among youth. Arch Pediatr Adolesc Med. 2006;160:18–24
   • View In Article
   • MEDLINE
   • CrossRef
5. Hersey JC, Niederdeppe J, Evans WD, et al. The theory of “truth”: how counterindustry media campaigns affect smoking behavior among teens. Health Psychol. 2005;24:22–31
   • View In Article
   • MEDLINE
   • CrossRef
6. Farrelly MC, Healton CG, Davis KC, Messeri P, Hersey JC, Haviland ML. Getting to the truth: evaluating national tobacco countermarketing campaigns. Am J Public Health. 2002;92:901–907
   • View In Article
   • MEDLINE
   • CrossRef
7. Emery S, Wakefield MA, Terry-McElrath Y, et al. Televised state-sponsored antitobacco advertising and youth smoking beliefs and behavior in the United States, 1999–2000. Arch Pediatr Adolesc Med. 2005;159:639–645
   • View In Article
   • MEDLINE
   • CrossRef
8. Longshore D, Ghosh-Dastidar B, Ellickson PL. National youth anti-drug media campaign and school-based drug prevention: evidence for a synergistic effect in ALERT plus. Addict Behav. 2006;31:496–508
   • View In Article
   • MEDLINE
   • CrossRef
9. Slater MD, Kelly KJ, Edwards RW, et al. Combining in-school and community-based media efforts: reducing marijuana and alcohol uptake among younger adolescents. Health Educ Res. 2006;21:157–167
   • View In Article
   • MEDLINE
   • CrossRef
10. McVey D, Stapleton J. Can anti-smoking television advertising affect smoking behaviour? (Controlled trial of the Health Education Authority for England's anti-smoking TV campaign). Tob Control. 2000;9:273–282
    • View In Article
    • MEDLINE
    • CrossRef
11. Snyder LB, Hamilton MA. A meta-analysis of U.S. health campaign effects on behavior: emphasize enforcement, exposure, and new information, and beware the secular trend. In:  Hornik RC editors. Public health communication: evidence for behavior change. Mahwah NJ: LEA; 2002;
    • View In Article
12. Huhman M, Potter LD, Wong FL, Banspach SW, Duke JC, Heitzler CD. Effects of a mass media campaign to increase physical activity among children: year-1 results of the VERB^™ campaign. Pediatrics. 2005;116:e277–e284
    • View In Article
    • CrossRef
13. Huhman ME, Potter LD, Duke JC, Judkins D, Heitzler CD, Wong FL. Evaluation of a national physical activity intervention for children: VERB^™ campaign 2002–2004. Am J Prev Med. 2007;32:38–43
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (868 KB)
    • CrossRef
14. Heitzler CD, Asbury LD, Kusner SL. Bringing “play” to life: the use of experiential marketing in the VERB^™ campaign. Am J Prev Med. 2008;34(6S):S188–S193
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (432 KB)
    • CrossRef
15. Bretthauer-Mueller R, Berkowitz JM, Thomas M, et al. Catalyzing community action within a national campaign: VERB^™ community and national partnerships. Am J Prev Med. 2008;34(6S):S210–S221
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (1730 KB)
    • CrossRef
16. Saunders RP, Pate RR, Felton G, et al. Development of questionnaires to measure psychosocial influences on children's physical activity. Prev Med. 1997;26:241–247
    • View In Article
    • MEDLINE
    • CrossRef
17. Potter LD, Judkins DR, Piesse A, Nolin MJ, Huhman M. Methodology of the outcome evaluation of the VERB^™ campaign. Am J Prev Med. 2008;34(6S):S230–S240
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (153 KB)
    • CrossRef
18. Rosenbaum O, Rubin D. The central role of the propensity score in observational studies for causal effects. Biometrika. 1983;70:41–55
    • View In Article
    • CrossRef
19. Kahn EB, Ramsey LT, Brownson RC, et al. The effectiveness of interventions to increase physical activity: a systematic review. Am J Prev Med. 2002;22:73–107
    • View In Article
    • Full Text
    • Full-Text PDF (34 KB)
    • CrossRef
20. Dearing JW. State of the art and state of the science of community organizing. In:  Thompson TL,  Dorsey AM,  Miller KI,  Parrott R editor. Handbook of health communication. Mahwah NJ: LEA; 2003;
    • View In Article
21. Welk G, Wickel EE, Peterson M, Heitzler C, Fulton JE, Potter LD. Reliability and validity of physical activity questions on the Youth Media Campaign Longitudinal Survey. Med Sci Sports Exerc. 2007;39:612–621
    • View In Article
    • MEDLINE
    • CrossRef