Outreach by the Numbers
North Carolina’s survey of residents’ stormwater awareness, perceptions, and behaviors
Public involvement and participation and public education and outreach on stormwater impacts make up just one-third of the EPA’s six National Pollutant Discharge Elimination System Phase II minimum management measures. Taken together, the two appear to create 10 times the confusion than the other four measures combined!
That’s not surprising—the average engineer knows as much about marketing as the average marketer knows about stormwater. However, stormwater management is a multidisciplinary field; both hard science and liberal arts approaches are required for success. Meshing the two can be a challenge, but North Carolina’s recent statewide stormwater survey provides a nexus where qualitative and quantitative thinkers can unite.
Social marketing is a term that encompasses everything from public events, article placement, and celebrity endorsements to community workshops, policy lobbyists, public service announcements, and giveaway items with hotline numbers. You’ve seen it before: think Click It or Ticket, Smokey Bear, and a host of anti-litter campaigns. Outreach and education perform a vital role in stormwater management that regulation and structures can never replace. As one anonymous regulator said, “Until we can permit the back end of a dog or a car window, voluntary action will be key in stormwater management.”
North Carolina’s first survey was created to measure residents’ awareness, perceptions, and behaviors related to polluted stormwater runoff. Funded by a Section 319 Clean Water Act grant from the EPA, the survey was administered by the North Carolina Division of Water Quality. The instrument was developed by East Carolina University’s Center for Survey Research in collaboration with the North Carolina Department of Environment and Natural Resources.
The survey is designed to be administered a total of three times. The instrument used in years two and three will feature questions about outreach campaigns conducted in the interim. Trends will be calculated on repeated questions to determine what changes, if any, occurred over time.
Knowing how much (or little) residents know about stormwater runoff tells campaign creators which messages to send where. When the audience is an entire state, finding subsets that share similarities like age, gender, education, and other qualities helps outreach professionals strategize.
Literature Review
Surveys measuring pre- and post-campaign effectiveness are an old tool, but given stormwater’s relatively new focus, their use in this field is recent. A review of the literature showed more local surveys than statewide efforts. Most were administered one time, so trend analysis was not possible. The areas examined varied, but water-quality and -quantity perception questions and personal behavior queries were a common theme.
Some general findings were constant regardless of where the survey was administered. Business, not people, is seen as the primary source of water pollution. South Carolina’s 2002 statewide phone survey showed that residents perceived industry as a larger threat to water quality than cities (University of South Carolina Institute for Public Policy and Policy Research 2002), and Tennessee’s 2003 survey showed residents citing agriculture, automotive fluids, and construction runoff as the biggest perceived threats (Gant and Daugherty 2003). A 1998 Colorado survey revealed that almost one-fifth of residents did not consider automotive fluids—typically an individual waste product—as a water-quality threat (ZumBrunnen 1998).
The tendency for individuals to hold business, industry, and large public enterprise responsible for water pollution is considered both a remnant of earlier outreach campaigns dating from more than 20 years ago and a common human tendency to blame negative events on external sources. Twenty years ago, business, industry, and large public facilities did represent the largest water-quality threats. Years of regulation applied to point-source water pollution dischargers, however, have substantially reduced the contaminants these entities produce. Now, the EPA’s research shows individual behaviors that create stormwater runoff represent the greatest threat to water quality. Obviously, the public still has some catching up to do.
Social psychologists have long noted that external impacts tend to be maximized over internal decisions when associated with negative outcomes. This tendency, known as “self-serving bias,” is the tendency for humans to take credit for success but to blame external causes for failure (Bernstein et al. 2003). When applied to social marketing efforts, the concept may play a role in many areas as diverse as weight loss (fast food), violent behaviors (media influence), and teen smoking (advertising). The probability of this concept playing a role with regard to individual perceptions of responsibility for stormwater is high. Regardless of the reason, the perceptions are reality in the minds of those who believe them. Attempts must be made to educate residents about the role they play and their responsibilities with regard to water quality.
Reported water-quality perceptions ranged considerably, however. Most Tennesseans rated their water quality as good, but city dwellers were more inclined to label it “fair” (Gant and Daugherty 2003) and express concern about the future. However, nearly half of Michigan’s Wayne County residents perceive their local river’s water quality as poor (Wayne County Department of Environment 1994) due to business and industrial waste (Ibid. p. 37).
Awareness of stormwater’s role regarding water quality remains lower than hoped. While South Carolina’s survey found more than half of its respondents considered stormwater to have a great impact on water quality (University of South Carolina Institute for Public Policy and Policy Research 2002), only a little more than a quarter of South Carolina residents know stormwater is not treated (Ibid. p. 3). However, fewer than 50% of Colorado’s residents understand that stormwater is not treated before entering local water bodies (ZumBrunnen 1998).
A range of individual behavior queries appeared on most statewide surveys. Both South Carolina and Tennessee collected data on yard fertilizing, pet waste, and grass clippings. The two differed on automotive fluids; Tennessee considered them standalone items (Gant and Daugherty 2003), whereas South Carolina grouped them with household hazardous waste (University of South Carolina Institute for Public Policy and Policy Research 2002). In Tennessee, more rural dwellers reported changing their own oil than their urban counterparts. Salt Lake County’s 2003 survey also showed that a quarter of residents changed their own oil, but 42% of them used commercial take-back programs (Dan Jones & Associates 2003).
In Tennessee, only one-fourth of car-washing residents reported washing vehicles in their driveways (Gant and Daugherty 2003), and almost one-fourth of Salt Lake County, UT, residents wash their vehicles at home, too. However, more than half (52%) of that group reported that they washed vehicles in their driveway (Dan Jones & Associates 2003).
Most surveys revealed few residents practice proper pet waste disposal. Close to one-third of South Carolina’s dog walkers reported they rarely or never pick up dog waste (University of South Carolina Institute for Public Policy and Policy Research 2002). Half of Tennesseans cite pet waste as a water pollutant (Gant and Daugherty 2003), but no data were gathered on disposal.
One-third of South Carolina residents fertilize their yards once per year, but rural residents report the lowest frequency (University of South Carolina Institute for Public Policy and Policy Research 2002). Less than one-third of the total population used soil tests, but females did so more than males. A full 60% of Tennessee residents use fertilizer regularly, with 25% doing so four times a year (Gant and Daugherty 2003). Only 25% of fertilizers reported using soil tests (Ibid. p. 3).
Methodology
North Carolina created a 31-item survey instrument in partnership with East Carolina University’s Center for Survey Research. In addition to stormwater awareness, water-quality perceptions, and behavioral data, respondents’ gender, age, income, education, and ethnicity were collected.
Data were collected between August 2005 and September 2005 using a sample of 11,200 telephone numbers purchased from a reputable survey sampling company. The same company grouped the telephone numbers into urban, suburban, and rural categories for coding.
East Carolina University hired and trained the interviewers to administer the survey using a computerized telephone-assisted interview software system. More than 11,000 calls were placed to capture 1,000 completed surveys, which yielded a statistical confidence level of 95% and a confidence interval of ± 3.1. After validity and reliability were established, data were analyzed using frequencies and cross-tabulations. For all statistical findings, significance levels were set at p = 0.05. Some statistically significant associations were found.
Although all attempts were made to obtain a representative sample, a chi-square goodness of fit test showed fewer males than females compared to 2000 Census data for the state. A significant difference was also noted between the sample and the population with regard to the race categories of African-American, Asian, White, Hispanic, and Other.
This survey used frequency counts and cross-tabulations as its primary statistical methodology. Frequency counts are like sports rankings. If every multiple-choice question held a contest among its possible answers, the highest scoring answer would be number one. Here, the respondents do the work, so you can learn, for example, the most popular used-oil-disposal method.
In survey analysis, comparing one subset’s answers to the whole group’s answers is called cross-tabulation. When statistical points of significance are seen, they act as signposts. For example, if you asked a group how often they got sunburned and their hair color, cross-tabulation might show more redheads and blondes burn when compared to the entire group. Consider the aggregate score as your background level and then look to see how subset scores compare to those levels.
Of course, if you are present, you can verify the hair color, but phone surveys rely on self-reported data. Social scientists have long been aware of the limitations inherent in self-reported data and have noted some general trends. Respondents may minimize negatively perceived behaviors (e.g., spanking, littering, speeding) and maximize positively perceived behaviors (e.g., charitable giving, quality time with children, volunteer hours donated). Despite the limitations, self-reported data from surveys still hold value. Focus groups yield richer data, but the logistics and costs involved are prohibitive and much of the data are still self-reported. Empirical methods do not suffer from self-report problems, but appointing watchers to record 1,000 state residents’ behaviors would create both cost and privacy problems.
Findings
Slightly more than one-third of North Carolina residents (37.6%) know that stormwater flows to the closest stream, lake, or river, and 13.2% believe it flows to drainage ponds, which may be a stormwater best management practice. However, 28.7% believe stormwater receives treatment at a special plant or at the sewer treatment plant. Fields and yards was the destination chosen by 9.8% of respondents, and 10.7% refused to answer the question.
The answer categories were combined into two categories: “treated” and “untreated.” Respondents who chose “the city’s regular sewer plant” or “a separate special sewer treatment plant” answers believe stormwater is treated. Those choosing one of three possible answers—“nearby fields and yards,” drainage pond,” or “closest stream, river or lake”—were placed into the “untreated” category.
Using these groupings, the “treated” category comprises 9% of men and 20% of women. In the “untreated” category, women made up 34% of the total and men composed 30%.
When asked to rate local water quality, the largest percentage (42.4%) described local streams, lakes, and rivers as “good,” with 39.4% choosing the “fair” rating and just 13.2% labeling them “excellent.” There was only one statistically significant finding with regard to water quality: Retirees were the only group significantly inclined to rate streams, lakes, and rivers as “good/excellent.”
When given a list of potential water-pollution sources, 93.4% of respondents thought recreational trash was “important/very important,” and 91.7% indicated that manufacturing wastewater was “important/very important.” In third place, 91% indicated that wastewater from sewer treatment plants was “important/very important.”
Other survey questions explored how residents cared for lawns, vehicles, and pets.
Lawn Care
North Carolina collected data on soil tests, fertilizer application, and yardwaste disposal. Most of North Carolina’s survey respondents (96%) reported they personally mow their yard. When asked how they disposed of grass clippings, the majority of this group (53.7%) reported leaving clippings in the yard. Mulching and composting yardwaste was chosen by 16.4% of respondents, but the second largest group of respondents (26%) reported they collect clippings for disposal in the garbage. A small but troubling 1.5% reported raking or blowing them into storm drains.
Grass clipping disposal is significantly influenced by education levels, but not in the intuitive sense. The only categories that did not report any storm drain disposal either attended or graduated from a vocational or technical program or held a postgraduate degree. All other educational levels reported some storm drain disposal, with high school graduates and respondents with some college reporting the most.
Because North Carolinians live in three distinct physiographic areas—the coast, the Piedmont, and the mountains—planting, fertilizing, and growing times vary statewide. Fewer than half (39.1%) of state residents reported they fertilize their own lawns. Of that group, 58% fertilize once a year or less. The next largest group (36%) fertilize two to three times per year, but a troubling 5% apply fertilizer monthly.
Annual household income was significantly associated with monthly fertilizer applications. Households earning more than $100,000 per year do so more than any other income level. Respondent age was also significantly associated; more 35- to 45-year-olds (12%) fertilize monthly than any other age group.
Because the best way to learn a yard’s fertilizer needs is to conduct a free soil test, respondents who stated they fertilized their own yard were also asked if anyone ever tested the soil. The majority (54%) did not use soil tests. Analysis showed soil testing was also significantly related to age. Respondents mostly likely to test their soil were aged 35 to 44 years, followed by those 45 to 54 years of age, with a third-place tie among 18- to 24-year-olds and 55- to 64-year-olds.
Vehicle Care
Three-fourths of respondents stated that they had a vehicle, but only 40% reported washing it at home. The majority (56%) wash in the driveway, but 41% wash atop gravel or grass.
Where respondents live significantly affects where they wash vehicles at home. Half of urban dwellers let soapy water drain into “the street or a driveway,” whereas 40% of suburban dwellers and 29% of rural dwellers report the same practice.
Less than one-fifth (16.7%) of vehicle owners change their own oil; most (76.9%) use commercial oil change facilities. Do-it-yourselfers disposed of oil in a variety of ways; no single method dominated. The most frequent destination (32.2%) was the garbage, but slightly more than one-fifth used an oil recycling facility. Two especially troubling findings cropped up: 22% of respondents dumped oil onto a designated part of their lawn, and 20.6% poured oil down storm drains.
Pet Waste
Research shows cat and raccoon wastes actually pose larger microbial threats than pet waste (Schueler and Holland 2000), but until these species are routinely guided on walks over impervious surfaces near or surrounding water bodies, the focus remains on man’s best friend.
Respondents who said they walked their pets were asked how often they picked up their pet’s waste. A significant relationship exists between proper waste disposal and dwelling area. While urban and suburban dwellers reported more pet-walking than their rural counterparts, respondents who stated they “rarely” or “never” picked up pet waste made up 47% of urban pet walkers, 49% of suburban pet walkers, and 59% of rural pet walkers.
Respondent age was significantly associated with pet-waste pickup. North Carolina’s youngest (18 to 24) and oldest (65 years and older) residents are most likely to report they “always” or “often” pick up pet waste. Gender was also found to be significantly associated; 35% of women stated they “always” or “often” pick up pet waste compared to 28% of the men.
Discussion
This survey’s findings create a roadmap for effective and efficient outreach and participation campaigns in North Carolina. Without understanding the link between personal behavior and water quality, humans have no motivation to make changes. True, knowledge in and of itself is insufficient to provoke change, but its absence pretty well guarantees that no change will occur. Despite years of public education, littering, drunk driving, and obesity continue. But declines in all three behaviors are attributed to a mix of social marketing, regulation, and best management practices such as biodegradable packaging, designated drivers, and low-fat menu choices. As with any complex system, changing a single input may or may not alter the output. Altering many inputs, however, vastly improves that probability. Outreach and education provide some of those inputs.
 |
Until residents know stormwater is not treated, they’ve little motivation to change behaviors. Because most North Carolinians fall into this group, awareness campaigns should precede or occur simultaneously with behavioral change messages. Traditionally, messages for widespread distribution utilize mass media like TV, cable, radio, and daily newspapers. In North Carolina, females are less aware than males, so outreach professionals may choose to augment mass media with other, more targeted outlets such as female-centric trade shows and women’s magazines.
For efficiency, specific behavioral messages should go to groups that need them most. North Carolina’s wealthiest homeowners apply the most fertilizer, so choosing media used heavily by this group avoids “waste.” But the survey also showed that most fertilizers of any income do not use soil tests. Here, messages about their availability should be targeted to all homeowners regardless of income. Broadcast media know a good deal about their audiences, so shows with high homeowner audiences can be identified through research. Gardening magazines, local home and garden shows, and point-of-purchase displays at garden supply stores would also reach this audience.
The yardwaste-disposal message will not be as easily delivered. The practice of improper disposal was associated with education, but this behavior and education levels did not rise together. Here, outreach would ideally be targeted at high school graduates and those with some college, but not toward those with vocational or technical education or those with postgraduate degrees. Mass media choices do not lend themselves as readily to targeting by education levels, so placing articles in school newsletters and campus events may be the best strategy.
Geography also plays a role. More urban and suburban dwellers wash vehicles on their driveways than rural residents, so outdoor billboards and bus shelters will direct messages to a specific area without paying the cost of delivery in lower-priority areas. Dwelling area was also associated with pet-waste pickup, so an outreach targeted to suburban and urban audiences makes common sense.
However, gender and age were also significantly related to pet-waste pickup. Females may be less aware of stormwater’s untreated status, but they are most likely to dispose of pet waste properly. Here, disposal messages can be efficiently directed toward urban and suburban men. Veterinarian partnerships in these areas would reach all pet owners, but the addition of kiosks at male-centric outlets like barbershops and local sporting events would weight the total effort. Because all adults aged 18 to 24 and 65 and older reported high proper disposal, focusing on middle-aged adults makes sense. The group is large enough that mass media outlets could be used if the media buy was placed with this target audience in mind.
Directing messages efficiently is only part of successful outreach; customizing message content is key. For example, we’re all aware of how gender, age, and other demographic attributes affect our motivations. Consider how large corporations use different approaches to sell the same product to different audiences. Once an outreach subgroup is identified, a wealth of data already exists about their needs, motivations, and values. Generally speaking, women may respond more readily to health messages than men, and while retirees on fixed incomes find cost a compelling point, it does not speak as loudly to those households earning more than $100,000. Once you have identified the audience, create a message that “speaks” to them and test it on a sample of your target audience. If it is well received, choose media that deliver the message most efficiently and you’ve just become a strategic communications planner.
It is important to realize social marketing’s limitations. As years of anti-littering, speeding, and teenage smoking campaigns prove, information is a necessary but insufficient ingredient to effect behavior change. The realization does not negate the need for social marketing efforts in this or any other area, but it does present a realistic framework to evaluate its value. Human behavior is a complex field that it still being explored. Social marketing messages are only one of many inputs influencing human behavior. The economy, natural phenomena, population changes, peer pressure, and a wealth of other factors can and do influence human behaviors, including those that impact water quality.
Financial constraints play a large role as well. The recommendations supplied here were “ideal” in that they did not consider cost constraints. While quite effective at delivering messages, mass media is expensive. North Carolina communities with Phase II permits are hard-pressed to supply funds for structural retrofits, much less a media campaign. Many have found that by working together, usually under the aegis of a council of government, they can afford more media of every type than they could if they chose to work alone. The result is that fewer messages are presented more often. As every teacher and most media buyers know, repetition aids retention. Seen from this perspective, the cost constraints may drive outreach professionals in the right direction, although for undesirable reasons. Unfortunately, the same constraints limit outreach globally. Every dollar added to this line item takes scarce dollars from other projects.
Advertisement
For these reasons, it is highly improbable any single campaign can provoke or prevent water-quality changes. Those caveats aside, leveraging social marketing with other inputs presents the best scenario for success with regard to behavior change. The work required to increase knowledge about stormwater and motivate water-protective behaviors exceeds the grasp of any one survey or campaign. Only harnessing the multitude of forces affecting water quality can do that. The breadth and depth of work required is daunting, but the goal makes the effort worthwhile.
For more details on the survey, visit http://www.ncstormwater.org/pdfs/stormwater_survey_12506.pdf to see the technical write-up, the instrument survey or its frequency count, and cross-tabulation appendices.
Author's Bio: Chrystal Bartlett is the stormwater awareness and outreach coordinator for the North Carolina Department of Environment and Natural Resources.
July-August 2006
Outreach by the Numbers
North Carolina’s survey of residents’ stormwater awareness, perceptions, and behaviors
Public involvement and participation and public education and outreach on stormwater impacts make up just one-third of the EPA’s six National Pollutant Discharge Elimination System Phase II minimum management measures. Taken together, the two appear to create 10 times the confusion than the other four measures combined! That’s not surprising—the average engineer knows as much about marketing as the average marketer knows about stormwater. However, stormwater management is a multidisciplinary field; both hard science and liberal arts approaches are required for success. Meshing the two can be a challenge, but North Carolina’s recent statewide stormwater survey provides a nexus where qualitative and quantitative thinkers can unite.
Social marketing is a term that encompasses everything from public events, article placement, and celebrity endorsements to community workshops, policy lobbyists, public service announcements, and giveaway items with hotline numbers. You’ve seen it before: think Click It or Ticket, Smokey Bear, and a host of anti-litter campaigns. Outreach and education perform a vital role in stormwater management that regulation and structures can never replace. As one anonymous regulator said, “Until we can permit the back end of a dog or a car window, voluntary action will be key in stormwater management.”
North Carolina’s first survey was created to measure residents’ awareness, perceptions, and behaviors related to polluted stormwater runoff. Funded by a Section 319 Clean Water Act grant from the EPA, the survey was administered by the North Carolina Division of Water Quality. The instrument was developed by East Carolina University’s Center for Survey Research in collaboration with the North Carolina Department of Environment and Natural Resources.
The survey is designed to be administered a total of three times. The instrument used in years two and three will feature questions about outreach campaigns conducted in the interim. Trends will be calculated on repeated questions to determine what changes, if any, occurred over time.
Knowing how much (or little) residents know about stormwater runoff tells campaign creators which messages to send where. When the audience is an entire state, finding subsets that share similarities like age, gender, education, and other qualities helps outreach professionals strategize.
Literature Review
Surveys measuring pre- and post-campaign effectiveness are an old tool, but given stormwater’s relatively new focus, their use in this field is recent. A review of the literature showed more local surveys than statewide efforts. Most were administered one time, so trend analysis was not possible. The areas examined varied, but water-quality and -quantity perception questions and personal behavior queries were a common theme.
Some general findings were constant regardless of where the survey was administered. Business, not people, is seen as the primary source of water pollution. South Carolina’s 2002 statewide phone survey showed that residents perceived industry as a larger threat to water quality than cities (University of South Carolina Institute for Public Policy and Policy Research 2002), and Tennessee’s 2003 survey showed residents citing agriculture, automotive fluids, and construction runoff as the biggest perceived threats (Gant and Daugherty 2003). A 1998 Colorado survey revealed that almost one-fifth of residents did not consider automotive fluids—typically an individual waste product—as a water-quality threat (ZumBrunnen 1998).
The tendency for individuals to hold business, industry, and large public enterprise responsible for water pollution is considered both a remnant of earlier outreach campaigns dating from more than 20 years ago and a common human tendency to blame negative events on external sources. Twenty years ago, business, industry, and large public facilities did represent the largest water-quality threats. Years of regulation applied to point-source water pollution dischargers, however, have substantially reduced the contaminants these entities produce. Now, the EPA’s research shows individual behaviors that create stormwater runoff represent the greatest threat to water quality. Obviously, the public still has some catching up to do.
Social psychologists have long noted that external impacts tend to be maximized over internal decisions when associated with negative outcomes. This tendency, known as “self-serving bias,” is the tendency for humans to take credit for success but to blame external causes for failure (Bernstein et al. 2003). When applied to social marketing efforts, the concept may play a role in many areas as diverse as weight loss (fast food), violent behaviors (media influence), and teen smoking (advertising). The probability of this concept playing a role with regard to individual perceptions of responsibility for stormwater is high. Regardless of the reason, the perceptions are reality in the minds of those who believe them. Attempts must be made to educate residents about the role they play and their responsibilities with regard to water quality.
Reported water-quality perceptions ranged considerably, however. Most Tennesseans rated their water quality as good, but city dwellers were more inclined to label it “fair” (Gant and Daugherty 2003) and express concern about the future. However, nearly half of Michigan’s Wayne County residents perceive their local river’s water quality as poor (Wayne County Department of Environment 1994) due to business and industrial waste (Ibid. p. 37).
Awareness of stormwater’s role regarding water quality remains lower than hoped. While South Carolina’s survey found more than half of its respondents considered stormwater to have a great impact on water quality (University of South Carolina Institute for Public Policy and Policy Research 2002), only a little more than a quarter of South Carolina residents know stormwater is not treated (Ibid. p. 3). However, fewer than 50% of Colorado’s residents understand that stormwater is not treated before entering local water bodies (ZumBrunnen 1998).
A range of individual behavior queries appeared on most statewide surveys. Both South Carolina and Tennessee collected data on yard fertilizing, pet waste, and grass clippings. The two differed on automotive fluids; Tennessee considered them standalone items (Gant and Daugherty 2003), whereas South Carolina grouped them with household hazardous waste (University of South Carolina Institute for Public Policy and Policy Research 2002). In Tennessee, more rural dwellers reported changing their own oil than their urban counterparts. Salt Lake County’s 2003 survey also showed that a quarter of residents changed their own oil, but 42% of them used commercial take-back programs (Dan Jones & Associates 2003).
In Tennessee, only one-fourth of car-washing residents reported washing vehicles in their driveways (Gant and Daugherty 2003), and almost one-fourth of Salt Lake County, UT, residents wash their vehicles at home, too. However, more than half (52%) of that group reported that they washed vehicles in their driveway (Dan Jones & Associates 2003).
Most surveys revealed few residents practice proper pet waste disposal. Close to one-third of South Carolina’s dog walkers reported they rarely or never pick up dog waste (University of South Carolina Institute for Public Policy and Policy Research 2002). Half of Tennesseans cite pet waste as a water pollutant (Gant and Daugherty 2003), but no data were gathered on disposal.
One-third of South Carolina residents fertilize their yards once per year, but rural residents report the lowest frequency (University of South Carolina Institute for Public Policy and Policy Research 2002). Less than one-third of the total population used soil tests, but females did so more than males. A full 60% of Tennessee residents use fertilizer regularly, with 25% doing so four times a year (Gant and Daugherty 2003). Only 25% of fertilizers reported using soil tests (Ibid. p. 3).
Methodology
North Carolina created a 31-item survey instrument in partnership with East Carolina University’s Center for Survey Research. In addition to stormwater awareness, water-quality perceptions, and behavioral data, respondents’ gender, age, income, education, and ethnicity were collected.
Data were collected between August 2005 and September 2005 using a sample of 11,200 telephone numbers purchased from a reputable survey sampling company. The same company grouped the telephone numbers into urban, suburban, and rural categories for coding.
East Carolina University hired and trained the interviewers to administer the survey using a computerized telephone-assisted interview software system. More than 11,000 calls were placed to capture 1,000 completed surveys, which yielded a statistical confidence level of 95% and a confidence interval of ± 3.1. After validity and reliability were established, data were analyzed using frequencies and cross-tabulations. For all statistical findings, significance levels were set at p = 0.05. Some statistically significant associations were found.
Although all attempts were made to obtain a representative sample, a chi-square goodness of fit test showed fewer males than females compared to 2000 Census data for the state. A significant difference was also noted between the sample and the population with regard to the race categories of African-American, Asian, White, Hispanic, and Other.
This survey used frequency counts and cross-tabulations as its primary statistical methodology. Frequency counts are like sports rankings. If every multiple-choice question held a contest among its possible answers, the highest scoring answer would be number one. Here, the respondents do the work, so you can learn, for example, the most popular used-oil-disposal method.
In survey analysis, comparing one subset’s answers to the whole group’s answers is called cross-tabulation. When statistical points of significance are seen, they act as signposts. For example, if you asked a group how often they got sunburned and their hair color, cross-tabulation might show more redheads and blondes burn when compared to the entire group. Consider the aggregate score as your background level and then look to see how subset scores compare to those levels.
Of course, if you are present, you can verify the hair color, but phone surveys rely on self-reported data. Social scientists have long been aware of the limitations inherent in self-reported data and have noted some general trends. Respondents may minimize negatively perceived behaviors (e.g., spanking, littering, speeding) and maximize positively perceived behaviors (e.g., charitable giving, quality time with children, volunteer hours donated). Despite the limitations, self-reported data from surveys still hold value. Focus groups yield richer data, but the logistics and costs involved are prohibitive and much of the data are still self-reported. Empirical methods do not suffer from self-report problems, but appointing watchers to record 1,000 state residents’ behaviors would create both cost and privacy problems.
Findings
Slightly more than one-third of North Carolina residents (37.6%) know that stormwater flows to the closest stream, lake, or river, and 13.2% believe it flows to drainage ponds, which may be a stormwater best management practice. However, 28.7% believe stormwater receives treatment at a special plant or at the sewer treatment plant. Fields and yards was the destination chosen by 9.8% of respondents, and 10.7% refused to answer the question.
The answer categories were combined into two categories: “treated” and “untreated.” Respondents who chose “the city’s regular sewer plant” or “a separate special sewer treatment plant” answers believe stormwater is treated. Those choosing one of three possible answers—“nearby fields and yards,” drainage pond,” or “closest stream, river or lake”—were placed into the “untreated” category.
Using these groupings, the “treated” category comprises 9% of men and 20% of women. In the “untreated” category, women made up 34% of the total and men composed 30%.
When asked to rate local water quality, the largest percentage (42.4%) described local streams, lakes, and rivers as “good,” with 39.4% choosing the “fair” rating and just 13.2% labeling them “excellent.” There was only one statistically significant finding with regard to water quality: Retirees were the only group significantly inclined to rate streams, lakes, and rivers as “good/excellent.”
When given a list of potential water-pollution sources, 93.4% of respondents thought recreational trash was “important/very important,” and 91.7% indicated that manufacturing wastewater was “important/very important.” In third place, 91% indicated that wastewater from sewer treatment plants was “important/very important.”
Other survey questions explored how residents cared for lawns, vehicles, and pets.
Lawn Care
North Carolina collected data on soil tests, fertilizer application, and yardwaste disposal. Most of North Carolina’s survey respondents (96%) reported they personally mow their yard. When asked how they disposed of grass clippings, the majority of this group (53.7%) reported leaving clippings in the yard. Mulching and composting yardwaste was chosen by 16.4% of respondents, but the second largest group of respondents (26%) reported they collect clippings for disposal in the garbage. A small but troubling 1.5% reported raking or blowing them into storm drains.
Grass clipping disposal is significantly influenced by education levels, but not in the intuitive sense. The only categories that did not report any storm drain disposal either attended or graduated from a vocational or technical program or held a postgraduate degree. All other educational levels reported some storm drain disposal, with high school graduates and respondents with some college reporting the most.
Because North Carolinians live in three distinct physiographic areas—the coast, the Piedmont, and the mountains—planting, fertilizing, and growing times vary statewide. Fewer than half (39.1%) of state residents reported they fertilize their own lawns. Of that group, 58% fertilize once a year or less. The next largest group (36%) fertilize two to three times per year, but a troubling 5% apply fertilizer monthly.
Annual household income was significantly associated with monthly fertilizer applications. Households earning more than $100,000 per year do so more than any other income level. Respondent age was also significantly associated; more 35- to 45-year-olds (12%) fertilize monthly than any other age group.
Because the best way to learn a yard’s fertilizer needs is to conduct a free soil test, respondents who stated they fertilized their own yard were also asked if anyone ever tested the soil. The majority (54%) did not use soil tests. Analysis showed soil testing was also significantly related to age. Respondents mostly likely to test their soil were aged 35 to 44 years, followed by those 45 to 54 years of age, with a third-place tie among 18- to 24-year-olds and 55- to 64-year-olds.
Vehicle Care
Three-fourths of respondents stated that they had a vehicle, but only 40% reported washing it at home. The majority (56%) wash in the driveway, but 41% wash atop gravel or grass.
Where respondents live significantly affects where they wash vehicles at home. Half of urban dwellers let soapy water drain into “the street or a driveway,” whereas 40% of suburban dwellers and 29% of rural dwellers report the same practice.
Less than one-fifth (16.7%) of vehicle owners change their own oil; most (76.9%) use commercial oil change facilities. Do-it-yourselfers disposed of oil in a variety of ways; no single method dominated. The most frequent destination (32.2%) was the garbage, but slightly more than one-fifth used an oil recycling facility. Two especially troubling findings cropped up: 22% of respondents dumped oil onto a designated part of their lawn, and 20.6% poured oil down storm drains.
Pet Waste
Research shows cat and raccoon wastes actually pose larger microbial threats than pet waste (Schueler and Holland 2000), but until these species are routinely guided on walks over impervious surfaces near or surrounding water bodies, the focus remains on man’s best friend.
Respondents who said they walked their pets were asked how often they picked up their pet’s waste. A significant relationship exists between proper waste disposal and dwelling area. While urban and suburban dwellers reported more pet-walking than their rural counterparts, respondents who stated they “rarely” or “never” picked up pet waste made up 47% of urban pet walkers, 49% of suburban pet walkers, and 59% of rural pet walkers.
Respondent age was significantly associated with pet-waste pickup. North Carolina’s youngest (18 to 24) and oldest (65 years and older) residents are most likely to report they “always” or “often” pick up pet waste. Gender was also found to be significantly associated; 35% of women stated they “always” or “often” pick up pet waste compared to 28% of the men.
Discussion
This survey’s findings create a roadmap for effective and efficient outreach and participation campaigns in North Carolina. Without understanding the link between personal behavior and water quality, humans have no motivation to make changes. True, knowledge in and of itself is insufficient to provoke change, but its absence pretty well guarantees that no change will occur. Despite years of public education, littering, drunk driving, and obesity continue. But declines in all three behaviors are attributed to a mix of social marketing, regulation, and best management practices such as biodegradable packaging, designated drivers, and low-fat menu choices. As with any complex system, changing a single input may or may not alter the output. Altering many inputs, however, vastly improves that probability. Outreach and education provide some of those inputs.
|
Until residents know stormwater is not treated, they’ve little motivation to change behaviors. Because most North Carolinians fall into this group, awareness campaigns should precede or occur simultaneously with behavioral change messages. Traditionally, messages for widespread distribution utilize mass media like TV, cable, radio, and daily newspapers. In North Carolina, females are less aware than males, so outreach professionals may choose to augment mass media with other, more targeted outlets such as female-centric trade shows and women’s magazines.
For efficiency, specific behavioral messages should go to groups that need them most. North Carolina’s wealthiest homeowners apply the most fertilizer, so choosing media used heavily by this group avoids “waste.” But the survey also showed that most fertilizers of any income do not use soil tests. Here, messages about their availability should be targeted to all homeowners regardless of income. Broadcast media know a good deal about their audiences, so shows with high homeowner audiences can be identified through research. Gardening magazines, local home and garden shows, and point-of-purchase displays at garden supply stores would also reach this audience.
The yardwaste-disposal message will not be as easily delivered. The practice of improper disposal was associated with education, but this behavior and education levels did not rise together. Here, outreach would ideally be targeted at high school graduates and those with some college, but not toward those with vocational or technical education or those with postgraduate degrees. Mass media choices do not lend themselves as readily to targeting by education levels, so placing articles in school newsletters and campus events may be the best strategy.
Geography also plays a role. More urban and suburban dwellers wash vehicles on their driveways than rural residents, so outdoor billboards and bus shelters will direct messages to a specific area without paying the cost of delivery in lower-priority areas. Dwelling area was also associated with pet-waste pickup, so an outreach targeted to suburban and urban audiences makes common sense.
However, gender and age were also significantly related to pet-waste pickup. Females may be less aware of stormwater’s untreated status, but they are most likely to dispose of pet waste properly. Here, disposal messages can be efficiently directed toward urban and suburban men. Veterinarian partnerships in these areas would reach all pet owners, but the addition of kiosks at male-centric outlets like barbershops and local sporting events would weight the total effort. Because all adults aged 18 to 24 and 65 and older reported high proper disposal, focusing on middle-aged adults makes sense. The group is large enough that mass media outlets could be used if the media buy was placed with this target audience in mind.
Directing messages efficiently is only part of successful outreach; customizing message content is key. For example, we’re all aware of how gender, age, and other demographic attributes affect our motivations. Consider how large corporations use different approaches to sell the same product to different audiences. Once an outreach subgroup is identified, a wealth of data already exists about their needs, motivations, and values. Generally speaking, women may respond more readily to health messages than men, and while retirees on fixed incomes find cost a compelling point, it does not speak as loudly to those households earning more than $100,000. Once you have identified the audience, create a message that “speaks” to them and test it on a sample of your target audience. If it is well received, choose media that deliver the message most efficiently and you’ve just become a strategic communications planner.
It is important to realize social marketing’s limitations. As years of anti-littering, speeding, and teenage smoking campaigns prove, information is a necessary but insufficient ingredient to effect behavior change. The realization does not negate the need for social marketing efforts in this or any other area, but it does present a realistic framework to evaluate its value. Human behavior is a complex field that it still being explored. Social marketing messages are only one of many inputs influencing human behavior. The economy, natural phenomena, population changes, peer pressure, and a wealth of other factors can and do influence human behaviors, including those that impact water quality.
Financial constraints play a large role as well. The recommendations supplied here were “ideal” in that they did not consider cost constraints. While quite effective at delivering messages, mass media is expensive. North Carolina communities with Phase II permits are hard-pressed to supply funds for structural retrofits, much less a media campaign. Many have found that by working together, usually under the aegis of a council of government, they can afford more media of every type than they could if they chose to work alone. The result is that fewer messages are presented more often. As every teacher and most media buyers know, repetition aids retention. Seen from this perspective, the cost constraints may drive outreach professionals in the right direction, although for undesirable reasons. Unfortunately, the same constraints limit outreach globally. Every dollar added to this line item takes scarce dollars from other projects.
For these reasons, it is highly improbable any single campaign can provoke or prevent water-quality changes. Those caveats aside, leveraging social marketing with other inputs presents the best scenario for success with regard to behavior change. The work required to increase knowledge about stormwater and motivate water-protective behaviors exceeds the grasp of any one survey or campaign. Only harnessing the multitude of forces affecting water quality can do that. The breadth and depth of work required is daunting, but the goal makes the effort worthwhile.
For more details on the survey, visit http://www.ncstormwater.org/pdfs/stormwater_survey_12506.pdf to see the technical write-up, the instrument survey or its frequency count, and cross-tabulation appendices.