The variables missing from the paid participants survey include demographics- story, tenure, exterior in the sun, maintenance – count of active leaks, behaviors – chemical storage, scented products, windows are open, frequently take a long shower, shoes are removed, and frequency of vacuuming.Table 4 presents the results from the negative binomial regression analysis of indoor characteristics, outdoor characteristics, and behavioral characteristics on the count of active health symptoms from the merged secondary dataset comprising responses from volunteer and paid participants. Table 4 presents the results of model 1 to model 5 of the negative binomial regression analysis. Model 1 describes the relationship between demographic characteristics and the count of active health symptoms, and this model accounts for 7% of the variance in active health symptoms. Model 2 describes the relationship between indoor characteristics and the count of active health symptoms, and this model explains 0% of the variance in active health symptoms after controlling for demographic characteristics. Model 3 describes the relationship between outdoor characteristics and the count of active health symptoms, and this model explains 3% of the variance in active health symptoms after controlling for demographic and indoor characteristics. Model 4 describes the relationship between maintenance behaviors and the count of active health symptoms, and this model explains 14% of the variance in active health symptoms after controlling for demographic, indoor, and outdoor characteristics. Model 5 describes the relationship between personal behaviors and the count of active health symptoms, and this model explains 0% of the variance in active health symptoms after controlling for demographic, indoor, outdoor characteristics, and maintenance behaviors. In general, the findings suggest that demographic such as home ownership, outdoor characteristics, such as living close to environmental hazards,microgreen grow rack and maintenance behaviors such as taking care of leaks, surface dust, and odors explain most of the variance in number of health symptoms among home occupants.
The most salient findings are related to demographic, outdoor characteristics, and maintenance behaviors. In terms of demographic, one of the variables in model 1 examines whether there is any difference in count of active health symptoms between volunteer and paid participants and it was found that volunteer participants were more likely to be suffering more health symptoms . Also, renters were more likely to experience more health symptoms than owners . In terms of outdoor characteristics , as expected, living close to hazards such as a highway , an industrial area , and farm are associated with more health symptoms. Noticeable dirt on sills , a result of environmental hazards is also associated with more health symptoms. In terms of maintenance behaviors, issues such as leaks , noticeable odor , and surface dust are associated with a greater number of health symptoms. In summary, in terms of demographics, renters are more likely to suffer more health symptoms than owners. In terms of outdoor characteristics, living close to highway, farm and industrial area are associated with more health symptoms. Environmental hazards can also affect occupants’ health through the presence of dirt on sills. In terms of maintenance behaviors, issues with leaks, noticeable odors, and surface dust are associated with more health symptoms. Finally, in terms of personal behaviors, smoking is associated with more health symptoms. A second regression analysis was performed to investigate whether maintenance behaviors moderate the effect of living near environmental hazards. Polychoric factor analyses were performed to obtain a negative maintenance behaviors scale and an environmental hazards scale because some of the items in the scale are on a dichotomous scale. The items on the negative maintenance behaviors scale are leaks, noticeable odors, surface dust, water stains, molds, pests , and the factor loadings of the variables ranged between 0.28 and 0.77. The items on the environmental hazards scale are proximity to airport, highway, industrial, coffee, dry cleaner, gas station, golf course, restaurant, and the factor loadings of the variables ranged between 0.28 and 0.86.
Negative maintenance is significantly associated with an increase in number of health symptoms which means that home occupants who tend to neglect maintenance issues such as leaks, odors, surface dust, water stains, molds, and pests are more likely to exhibit more health symptoms . Environmental hazards are also significantly associated with an increase in number of health symptoms which means that home occupants who live near to environmental hazards are more likely to exhibit more health symptoms. An interaction between environmental hazards and negative maintenance was carried out in model 6 of the regression model to find out if maintenance might mitigate the effect of living near environmental hazards. However, the interaction was insignificant which suggests that among home occupants who live near environmental hazards, there is no difference in number of health symptoms among those who engage in frequent maintenance and those who do not.A robustness check was performed to compare the responses of the paid participants to those of the volunteer participants are comparable. Comparison of analyses results as a check of robustness have been documented in previous studies . A notable difference is that paid participants reported experiencing more health symptoms than volunteer participants which suggests that there might be other characteristics that differentiate them. Despite the difference in number of health symptoms experienced by both groups of participants, the variance in count of active health symptoms as explained by indoor, outdoor, and behavioral characteristics are consistent across both datasets. In both datasets, maintenance behaviors explained the greatest variance in count of active health symptoms, 9% in the dataset with responses from paid participants, and 7% in the dataset with responses from volunteer participants. Other than maintenance behaviors, outdoor characteristics also explained a sizable variability in count of active health symptoms, 6% in the dataset with responses from paid participants, and 3% in the dataset with responses from volunteer participants, followed by demographics which explained 3% in the dataset with responses from paid participants,ebb and flow flood table and 2% in the dataset with responses from volunteer participants.
The dataset with responses from the volunteer participants contains more variables than the dataset with responses from the paid participants. The additional variables in the volunteer participants dataset provide useful insights into the effects of beneficial and harmful behaviors such as the presence of carpet, frequency of vacuuming, and the use of air purifier on health symptoms. A deeper investigation into the role of these variables in mitigating or worsening indoor and outdoor characteristics, and thereby health symptoms was explored through structural equation modeling and is presented in the next section. One of the limitations of this study is that volunteer participants were motivated to participate because they probably were more likely to experience issues in their homes and/or experienced health symptoms, and this might affect the generalizability of the findings. To mitigate the effect of an unrepresentative sample, similar analyses were conducted with both the volunteer participants and the more representative paid participants . The variability in health outcomes explained by indoor, outdoor, and behavioral characteristics were similar across both groups of participants, and in both groups, demographics, outdoor characteristics, and maintenance behaviors were found to be the greatest contributors to health outcomes. While it is not possible to account fully for the bias that might occur with the volunteer participants, the study findings can still benefit individuals, especially those with health conditions who are interested in behaviors that can mitigate the impact of poor indoor air quality. Future studies should be conducted with a more representative population, thus avoiding the problem of self-selection to examine if the contribution of outdoor characteristics and maintenance behaviors to health symptoms is similar to what was found in this study. Subsequent studies should also investigate if the effect of living near environmental hazards and maintenance issues such as leaks are just as deleterious among healthy individuals, and whether the moderating behaviors highlighted in the study findings can mitigate their negative effect on health. The recruitment of volunteer participants who are likely to experience more issues in their homes and/or experienced more health issues are also likely to influence some of the study findings, for instance the use of air purifier were found to result in more health symptoms. It could be the case that participants who had more health symptoms were more likely to use air purifier. Another limitation of this study is that the data from this study were obtained from participants’ self-reported surveys which could result in issues such as social desirability, difficulty with retrieval, and judgment with a self-reported survey .
Social desirability occurs when participants are inclined to respond to survey questions so that they will be viewed in a favorable light, for instance downplaying the negative issues in their homes or the number of health symptoms that they experienced. Problem with retrieval occurs when participants must recall instances, for example, participants might not be accurate in their recall of the frequency of meal preparation. Problem with judgment occurs when the participants face an issue matching the recalled instances to the scale context. For instance, the options in the frequency of maintenance question were deferred maintenance, somewhat maintained, and highly maintained; however, participants might have a different interpretation of what constitutes highly maintained and somewhat maintained and recalled instances might map differently across participants. The final limitation has to do with the fact that all the models in the regression analysis could only explain 24% of the variance in active health symptoms. The moderate variance explained by the models is not surprising as health outcomes are affected by a multitude of factors, beyond what was covered in the survey. Other than the factors described in the survey, health outcomes can also be affected by diet and exercise, use of alcohol and drugs, quality of clinical care, education attainment, employment status, family and social support, and community safety . Land use change is one of the greatest threats to wildlife worldwide—globally, it can remove and alter habitat, or disrupt wildlife interactions . A major challenge for conservation involves navigating the negative environmental repercussions of land use change alongside the needs for human agriculture and development . This means that studying land use change fundamentally engages the role of humans within ecological systems and processes . Research has increasingly focused on human impacts on surrounding ecosystems, revealing complex interactions and consequences . However, mechanistic understanding, universal rules, or consistent predictions are difficult to define, and more context-based research is needed, especially in systems early in the process of land use transition. Cannabis agriculture provides an ideal opportunity to study ecological outcomes of land use change in a rural and rapidly changing landscape. To understand why, it is important to start with the recent history of cannabis cultivation in the western US. For decades, cannabis was grown illegally in rural areas of California, Oregon, and Washington as part of the back-to-the-land movement . These were remote areas that allowed counter-culture communities to reinvent themselves, but which also happened to host some of the nation’s highest biodiversity . The industry remained surreptitious and small-scale for many years, while ongoing law enforcement and the US “war on drugs” tried unsuccessfully to eliminate the practice . Then, the ground shifted with recreational legalization. Oregon passed recreational Adult Use cannabis legalization in the fall of 2015, and California followed suit a year later, riding a wave of recreational legalization measures that eventually passed across 19 states in the US. Very rapidly, this policy change initiated land use development for cannabis , first in areas with a history of cultivation, and later, into new regions. This shift in development was accompanied by subtle shifts in motives and philosophy behind cannabis cultivation – as one of the farmers I interviewed for Chapter 2 put it, “The quest for the all mighty dollar got in the way of the spiritual cycle of the plant.” Along with these rapid changes came calls of concern for potential environmental impacts . However, the illicit history of cannabis meant that there was very little existing research on cannabis-environment interactions, and many gaps in baseline data . To address this brewing conservation crisis, I focused my dissertation on the ecological outcomes of cannabis legalization. I was specifically interested in studying private land cannabis development in rural areas with a history of pre-legalization cultivation . In these regions, legalization has spurred major private land development for cannabis alongside high biodiversity and few other crop based agricultural land uses.