Abstract
Introduction:
Understanding the multi-dimensional risk beliefs of agricultural audiences allows risk communicators and educators to target those beliefs to improve practices. This study was designed to assess pesticide risk beliefs among English-speaking farmers and Spanish-speaking farmworkers and to compare their beliefs.
Methods:
The Pesticide Risk Beliefs Inventory (PRiBI) is a 19-item quantitative instrument used to assess alignment of risk beliefs with those beliefs of experts in the field. A higher score on the PRiBI relates to agreement with expert beliefs regarding pesticide risk.
Results and Discussion:
Farmers’ and farmworkers’ scores were significantly different on items corresponding to use of physical properties to determine risk and specific adverse health outcomes associated with pesticide exposure.
Conclusions:
With an understanding that farmworkers rely on physical properties of pesticides to assess risk, educators and farmers can encourage more reliable ways to assess pesticide hazards.
Introduction
Previous studies have demonstrated that agricultural workers’ beliefs about pesticide risk influence their pesticide practices, which in turn impact their potential exposure [1]. For example, farmers who are more accepting of risk are less likely to use protective equipment and more likely to apply pesticides in close proximity to their homes and drinking water wells [2]. These linkages between risk beliefs, practices, and exposure suggest that an understanding of beliefs may help educators target education and training to shape beliefs, improve practices, and reduce adverse outcomes [3,4].
In the United States, agricultural workers who may be exposed to pesticides include both farmers—individuals who own farm land and manage crop production, and farmworkers—individuals sometimes hired by farmers to supply the manual labor necessary to cultivate and harvest crops. Many farmers have a license to apply pesticides in the production of an agricultural commodity on land they own or lease. To maintain a pesticide applicator’s license, the United States Environmental Protection Agency requires recertification at least every 5 years to ensure competency through either continuing education or testing [5].
While farmworkers may not apply pesticides themselves, pesticide exposure is a significant hazard for farmworkers in both their working and living environments [6, p.29–63]. The pesticide training farmworkers receive depends on the tasks in which they engage on the farm (e.g., applying a pesticide under the supervision of a licensed applicator, entering a field during a restricted entry interval, or harvesting).
In the last decade, a number of studies have been conducted with members of the agricultural workforce in countries outside of the United States. These studies have tended to focus on workers’ pesticide knowledge and the relationship between knowledge and practice (e.g., 7–9). While they confirm the importance of beliefs in shaping practice, recent studies do not afford an understanding of multi-faceted risk beliefs that could be targeted for intervention to improve practice. The present study was designed to characterize pesticide risk beliefs among farmers and farmworkers in North Carolina and to compare their beliefs. These beliefs included the usefulness of physical and chemical properties in determining risk and the association of risk with pesticide routes of entry into the body and with adverse health outcomes resulting from pesticide exposure.
Methods
Sample/setting
To maintain their licenses, farmers in North Carolina must participate in continuing education courses offered at county Cooperative Extension centers. The farmers in this study responded to the Pesticide Risk Belief Inventory (PRiBI) and demographic items during continuing education courses. Participation was voluntary and independent of applicators’ receiving credit for the continuing education courses. A total of 131 farmers responded to the questionnaire. See Table 1 for demographic characteristics of the participating farmers.
Table 1.
Participant demographics by worker group
| Overall n=203 |
Farmers n=131* |
Farmworkers n=72 |
test statistic | p-value | |
|---|---|---|---|---|---|
| Age | 47.62 (17.66) | 56.40 (14.50) | 32.75 (11.54) | t=11.81 | <.001 |
| Gender | |||||
| Female | 9 (4.48%) | 6 (4.65%) | 3 (4.17%) | χ2=0.03 | .87 |
| Male | 192 (95.52%) | 123 (95.35%) | 69 (95.83%) | ||
| Race/ethnicity | χ2=201.00 | <.001 | |||
| African-American | 5 (2.49%) | 5 (3.88%) | 0 (0%) | ||
| Native American | 37 (18.41%) | 37 (28.68%) | 0 (0%) | ||
| Pacific Islander | 1 (0.50%) | 1 (0.78%) | 0 (0%) | ||
| Latino | 72 (35.82%) | 0 (0%) | 72 (100.00%) | ||
| White | 86 (42.79%) | 86 (66.67%) | 0 (0%) | ||
| Country of origin | χ2=194.00 | <.001 | |||
| France | 1 (0.52%) | 1 (0.82%) | 0 (0%) | ||
| Honduras | 2 (1.03%) | 0 (0%) | 2 (2.78%) | ||
| Mexico | 70 (36.08%) | 0 (0%) | 70 (97.22%) | ||
| United States | 121 (62.37%) | 121 (99.18%) | 0 (0%) |
Note. Means and standard deviations reported for continuous age. N sizes and percentages reported for categorical variables.
For the farmer group, total responses for each characteristic do not equal 131 where one or more participants did not respond to the corresponding item.
The farmworkers in this study responded to the PRiBI and demographic items as part of a verbally-administered questionnaire in their residences. The farmworkers who participated in this study were recruited as part of a larger study in North Carolina focused on pesticide protective behaviors [10]. Individual participation was voluntary. A total of 72 farmworkers were administered and responded to the questionnaire. See Table 1 for demographic characteristics of the participating farmworkers. The procedures described in this paper were approved by the Duke University Institutional Review Board.
Measures
The Pesticide Risk Beliefs Inventory (PRiBI), initially validated with pesticide educators [11], contains 19 Likert-type items with six-point scales. Ten items are reverse coded. For the purpose of scoring and analyzing the PRiBI data, the response “strongly disagree” corresponds to a score of 1, “disagree” corresponds to 2, “somewhat disagree” corresponds to 3, “somewhat agree” corresponds to 4, “agree” corresponds to 5, and “strongly agree” corresponds to 6. For reverse-coded items, “strongly disagree” corresponds to a numerical score of 6 and so forth. A score of 4 or higher on the PRiBI relates to agreement with an expert-like belief regarding pesticide risk.
When adapted for administration to farmworkers in this study, the same 19 Likert-type PRiBI items were translated into Spanish and administered with four-point scales. Two independent translations of the the PRiBI and demographic items were completed and the versions compared. Minor differences (i.e., use of slang and dialectical language) in the two versions were discussed and resolved by the study team. The decision to utilize a four-point scale was made based on literature suggesting that in low-literacy immigrant Latino populations, a larger number of response items may not be well understood and respondents may tend toward excessive use of the endpoints of the scales [12–14]. During PRiBI administration, farmworkers were presented a visual color aid with gradations of blue and a number assigned to each hue to contextualize the intensity of their responses. For the first 10 items, the lightest hue corresponded to a score of 1 and a response of “strongly disagree” and the darkest hue corresponded to a score of 4 and a response of “strongly agree.” For the remaining 9 items, the lightest hue corresponded to a score of 1 and a response of “strongly agree” and the darkest hue corresponded to a score of 4 and a response of “strongly disagree.” Scores for items 11–19 were then reversed such that a score of 1 corresponded to “strongly disagree” and a score of 4 corresponded to “strongly agree”, in alignment with the first 10 items. . Demographic items included age, gender, race/ethnicity, and country of origin.
Data collection
Farmers independently responded to a paper-based questionnaire containing PRiBI and demographic items. The questionnaire was available and administered to farmers in English only. A native Spanish speaker whose parents were farmworkers and who grew up close to the farms in the study was trained to verbally administer the PRiBI and demographic items to the farmworkers. All farmworkers responded in Spanish.
Data analysis
SAS Version 9.4 was used to conduct all statistical analyses. Descriptive statistics were computed for demographics. Frequencies were calculated for categorical variables (gender, race/ethnicity, and country of origin), and sample means and standard deviations were calculated for the continuous variable (age). We also tested bivariate relationships between demographics and worker group, using chi-square tests of independence for categorical variables and a Pearson correlation coefficient for the continuous variable.
In order to compare farmworker and farmer responses to the PRiBI items, we expanded the 4-point scale used by farmworker respondents to a 6-point scale as was used by farmer respondents. Farmworker scores for “agree” shifted from 3 to 5, and for “strongly agree” shifted from 4 to 6. No modifications were made for farmworker responses for “strongly disagree” (1) or “disagree” (2). To assess relationships between the 19 PRiBI items and worker group, we used independent-samples t-tests. For these tests, we used a Bonferroni-adjusted p-value of .003 to reduce Type I error associated with multiple testing for the 19 individual items.
Results and Discussion
The average age of farmers (n=122) was 56.4 years (SD=14.5, range 21–83 years) and of farmworkers was 32.8 years (SD=11.5, range 18–68 years). Both worker groups were overwhelmingly male. Farmers were predominantly white and from the United States. Farmworkers were predominantly Latino and from Mexico. Table 1 indicates age to be significantly related to worker group (t=11.81, p<.001), with farmers being older than farmworkers. In addition, worker group was strongly related to race/ethnicity (χ2= 201.00, p<.001) and country of origin (χ2=194.00, p<.001), in that all farmworkers were Latinos from Mexico or Honduras. Individual worker groups were largely homogenous but differed from each other with regard to age, race/ethnicity, and country of origin.
Among farmworkers, the response rate was 100% for each PRiBI item. Among farmers, the response rates for individual items ranged from 88.5% to 98.5%. Table 2 presents mean scores on PRiBI items by worker group. Findings indicated that, contrary to expert beliefs, farmworkers agreed more strongly than farmers with the statements that they could determine if a pesticide was dangerous by its smell (t= 14.02, p< .001); by seeing whether it is a powder, liquid, or granule (t=14.05, p<.001); by its color (t=10.26, p<.001); and by its taste (t=8.03, p<.001). Farmworkers were more worried than farmers about having a recurrent problem with their skin (t=−4.05, p<.001), about having to go to the emergency room (t=−7.28, p<.001), and about losing their ability to have children (t=−4.87, p<.001). Responses on the remaining PRiBI items, as well as total PRiBI score, did not differ significantly by worker group.
Table 2.
PRiBI items by worker group
| Overall n=203 |
Farmers n=131 |
Farmworkers n=72 |
t- value |
p- value |
|
|---|---|---|---|---|---|
| 1. I can determine if a pesticide is dangerous by its smell.* | 3.99 (1.88) | 4.99 (1.39) | 2.24 (1.23) |
14.02 | <.001 |
| 2. When I am working with a pesticide, I am worried about having the pesticide enter my body when I breathe. | 4.87 (1.29) | 4.97 (1.21) | 4.68 (1.41) |
1.52 | .13 |
| 3. I can determine if a pesticide is dangerous by reading its chemical label. | 5.21 (1.09) | 5.29 (1.07) | 5.07 (1.12) |
1.36 | .18 |
| 4. When I am working with a pesticide, I am not worried about getting cancer in the future.* | 4.47 (1.51) | 4.36 (1.51) | 4.67 (1.50) |
−1.40 | .16 |
| 5. I can determine if a pesticide is dangerous by seeing whether it is a powder, liquid, or granule.* | 4.18 (1.80) | 5.13 (1.20) | 2.49 (1.40) |
14.05 | <.001 |
| 6. When I am working with a pesticide, I am worried about having a recurrent problem with my skin. | 4.35 (1.43) | 4.05 (1.42) | 4.88 (1.29) |
−4.05 | <.001 |
| 7. When I am working with a pesticide, I am not concerned about covering my nose.* | 4.74 (1.32) | 4.62 (1.44) | 4.94 (1.06) |
−1.81 | .07 |
| 8. When I am working with a pesticide, I am not worried about having the pesticide enter my body through my skin.* | 4.91 (1.35) | 4.89 (1.36) | 4.96 (1.35) |
−0.34 | .73 |
| 9. When I am working with a pesticide, I am worried about having to go to the emergency room. | 3.91 (1.52) | 3.38 (1.43) | 4.83 (1.20) |
−7.28 | <.001 |
| 10. I can determine if a pesticide is dangerous by its color.* | 4.57 (1.60) | 5.35 (0.92) | 3.21 (1.63) |
10.26 | <.001 |
| 11. When I am working with a pesticide, I am concerned about covering my skin. | 5.06 (0.98) | 5.09 (0.87) | 5.01 (1.14) |
0.47 | .64 |
| 12. I can determine if a pesticide is dangerous by knowing the (pesticide’s) family of chemicals. | 3.93 (1.50) | 3.79 (1.46) | 4.17 (1.55) |
−1.67 | .10 |
| 13. I can determine if a pesticide is dangerous by knowing its ingredients. | 4.29 (1.45) | 4.26 (1.37) | 4.33 (1.59) |
−0.34 | .74 |
| 14. When I am working with a pesticide, I am not worried about having the pesticide enter my body through my eyes.* | 4.69 (1.45) | 4.75 (1.45) | 4.60 (1.46) |
0.71 | .48 |
| 15. When I am working with a pesticide, I am worried about losing my ability to have children. | 3.76 (1.65) | 3.32 (1.48) | 4.46 (1.68) |
−4.87 | <.001 |
| 16. When I am working with a pesticide, I am worried about having the pesticide enter my body when I eat or drink. | 4.47 (1.40) | 4.38 (1.38) | 4.61 (1.43) |
−1.10 | .27 |
| 17. When I am working with a pesticide, I am not worried about having difficulty breathing.* | 4.30 (1.55) | 4.32 (1.46) | 4.26 (1.70) |
0.24 | .81 |
| 18. When I am working with a pesticide, I am not worried about being poisoned.* | 4.64 (1.47) | 4.58 (1.42) | 4.76 (1.55) |
−0.87 | .39 |
| 19. I can determine if a pesticide is dangerous by its taste.* | 4.92 (1.53) | 5.56 (0.97) | 3.82 (1.69) |
8.03 | <.001 |
| Total Score | 82.95 (14.46) | 83.48 (15.99) | 81.99 (11.18) | 0.78 | .44 |
Note. 1=strongly disagree and 6=strongly agree. For reverse coded items (indicated by *), 1=strongly agree and 6=strongly disagree. Bonferroni corrected alpha level: p=.003.
Of the seven items for which there were significant differences by worker group, three items were related to adverse health outcomes of pesticide exposure and four items were related to determining a pesticide’s danger by its physical properties. These findings show that this group of farmworkers was more concerned about specific adverse health outcomes than the corresponding farmer group. Additionally, the farmworkers believed that physical properties of pesticides were more useful in determining risk.
Conclusions
The differences observed in risk beliefs by worker group have implications for both training and pesticide risk communication on the farm. Pesticide safety educators benefit from understanding their audiences’ risk beliefs so that the educators can tailor training to ultimately improve practices [3,4]. Educators of farmworkers should consider explicitly addressing the limitations of relying on the physical properties of pesticides to assess risk. Recognizing farmworkers’ concerns about specific health adverse health outcomes associated with pesticide exposure, farmworker educators might frame pesticide protective behaviors as a way to reduce these health outcomes.
While training provided by pesticide safety educators is a critical component of promoting protective behavior and risk mitigation practices, the communication that occurs daily on farms between farmers and their farmworker employees provides an opportunity for ongoing dialogue about pesticide risk. When farmers are cognizant that farmworkers rely more heavily on physical properties of pesticides to determine risk, they can emphasize warnings when it is particularly problematic to rely on these physical properties and encourage more reliable ways to assess pesticide hazards.
Funding Details:
This work was supported by the National Institute of Nursing Research of the National Institutes of Health (T32NR013456, T32NR007091), the American Cancer Society (DSCNR-13-276-03), and the Southeast Center for Agricultural Health & Injury Prevention.
Footnotes
Disclosure Statement: The authors declare no conflict of interest.
References
- 1.Damalas CA, Koutroubas SD. Farmers’ behaviour in pesticide use: A key concept for improving environmental safety. Curr Opin Environ Sci Health. 2018;4:27–30. [Google Scholar]
- 2.DellaValle CT, Hoppin JA, Hines CJ, et al. Risk-accepting personality and personal protective equipment use within the Agricultural Health Study. J Agromedicine. 2012;17(3):264–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Damalas CA, Koutroubas SD. Farmers’ Training on Pesticide Use Is Associated with Elevated Safety Behavior. Toxics. 2017Aug 22;5(3). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Schreinemachers P, Chen H, Nguyen TTL, et al. Too much to handle? Pesticide dependence of smallholder vegetable farmers in Southeast Asia. Sci Total Environ. 2017Sep1;593–594:470–477. [DOI] [PubMed] [Google Scholar]
- 5.Federalregster.gov [Internet]. Washington (DC): Federal Register; [cited2020 Feb 20]. Available from: https://www.federalregister.gov/documents/2017/01/04/2016-30332/pesticides-certification-of-pesticide-applicators [Google Scholar]
- 6.Donham KJ, Thelin A Special Risk Populations in Agricultural Communities. In Agricultural Medicine: Occupational and Environmental Health for the Health Professions; Donham KJ, Thelin A, Eds.; Blackwell: Ames, IA, USA, 2006; p. 29–63. [Google Scholar]
- 7.Taghdisi MH, Dehdari T, Khalili F, et al. Knowledge and Practices of Safe Use of Pesticides among a Group of Farmers in Northern Iran. Int J Occup Environ Med. 2019Apr 1; 10(2): 66–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Wang W, He R, Gong H, et al. Gender differences in pesticide use knowledge, risk awareness and practices in Chinese farmers. Sci Total Environ. 2017Jul15;590–591:22–28. [DOI] [PubMed] [Google Scholar]
- 9.Ricco M, Vezzosi L, Gualerzi G Health and safety of pesticide applicators in a high income agricultural setting: a knowledge, attitude, practice, and toxicity study from North-Eastern Italy. J Prev Med Hyg. 2018Sep 28;59(3):E200–E211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Walton AL, LePrevost C, Wong B, Linnan L, Sanchez-Birkhead A, Mooney K. Observed and self-reported pesticide protective behaviors of Latino migrant and seasonal farmworkers. Environ Res. 2016;147:275–283. doi: 10.1016/j.envres.2016.02.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.LePrevost CE, Blanchard MR, Cope WG. The Pesticide Risk Beliefs Inventory: a quantitative instrument for the assessment of beliefs about pesticide risks. Int J Environ Res Public Health. 2011;8(6):1923–1935. doi: 10.3390/ijerph8061923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.D’Alonzo KT. Evaluation and revision of questionnaires for use among low-literacy immigrant Latinos. Rev Lat Am Enfermagem. 2011Sep-Oct;19(5):1255–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bernal H, Wooley S, Schensul JJ .The challenges of using Likert-type scales with low-literate ethnic populations. Nurs Res. 1997;46(3):179–181. [DOI] [PubMed] [Google Scholar]
- 14.McQuiston C, Larson K, Parrado EA, et al. AIDS knowledge and measurement considerations with unacculturated Latinos. West J Nurs Res. 2002June;24(4):354–72. [DOI] [PubMed] [Google Scholar]