Efficient cookstoves are popular solutions in international energy development for their presence in multiple Sustainable Development Goals. However, adoption of these devices remains a challenge. Among the barriers known to limit adoption is effective household communication, yet rarely are all stakeholders purposefully engaged in activities that support uptake and sustained use. The findings of two studies conducted in rural Namibia indicate that youth-oriented sustainability education is an effective communication tool in cooking energy development, and has the potential to shift attitudes and behaviours, within households and across communities. Households with children who participated in sustainability education programming were significantly more likely to prefer efficient fuels and be less committed to biomass cookstoves than similar families without such children. Further, a longitudinal study of children who attended the programming indicates that their preferences about common cooking fuels shifted toward sustainable ones, and that these changes persist over time.
Article
Electric cookstove adoption and energy attitudes in rural Namibia: The role of youth and sustainability education in advancing household energy
https://doi.org/10.21203/rs.3.rs-1422932/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted 21 Mar, 2022
You are reading this latest preprint version
Taken together, the findings from these 2 quasi-experimental studies provide compelling evidence for the efficacy of youth-centric sustainability education in advancing sustainable cooking and household energy within the home and across communities. Households with a member exposed to local, youth-oriented sustainability education programming were found to have significantly more positive attitudes regarding solar energy and cookers and preferred more efficient fuels. For those that can more easily afford electricity, the sustainability education programming appears to support an increase in electric, and a decrease in traditional, stove commitments. The data also suggest that youth sustainability education positively impacts the household’s attitudes and knowledge in the absence of an affordable behavioural change.
Findings from Stampriet highlight the agentive capacity of youth. Households with children but without exposure to NaDEET’s programming, preferred traditional cooking energy and stoves and were significantly less positive about electric stoves and other efficient energy sources than households with NaDEET experience. Similarly, the significant gains in sustainable energy knowledge and attitudes made by NaDEET’s 2019 youth participants indicates that sustainability education programming has the potential to shift perceptions and understanding and demonstrates that youth do exert some influence over the attitudes and behaviours within their homes.
Moreover, attitudinal differences between non-NaDEET households in Stampriet and the control town suggest that NaDEET’s youth programming has a positive impact on local communities. Students who attend NaDEET are not given stoves nor talking points for their return home, and yet improved energy attitudes and electric stove commitments occurred. This indicates that youth education contributes to a home environment in which attitudinal and behavioural changes are supported, a key pillar of Social Cognitive Theory. While this has implications for supporting development work committed to realizing the Sustainable Development Goals (SDGs) generally, the findings described herein may be of particular interest to the cookstove community, because of the behaviour dependent nature of cookstove uptake and sustained use.
While youth participants exhibited positive gains in attitudes and knowledge after just one week, it is not known how quickly these beliefs are transmitted through the household or broader community, though it is clear from the evidence presented here and elsewhere (blinded reference) that both do occur. To date, little research has been done on the role that this type of education has in shaping families’ energy beliefs and consumption patterns, particularly in resource-limited settings. Locally led and contextually appropriate youth education is a potential mechanism for supporting cookstove adoption efforts.
Youth are not only household energy stakeholders, but also “critical agents of change” for their potential to act on a large scale and because they will be most impacted by a failure to realize the SDGs50. Today’s youth will soon be in positions of energy decision-making for themselves and for their own households. Investing in education now has the potential to advance the long-term goals of both the SDGs and those of the cookstove community.
Community Survey Design
Household interviews in Stampriet and Gibeon, consisted of survey questionnaires and open-ended interview prompts, and were conducted in March and April 2019. A stratified design with random walk was utilized. Census enumeration maps available online from Digital Namibia and Google Earth images were used to segment each community into strata.
Based on aerial images available, each community was divided into 12-16 strata along naturally occurring divisions such as streets, dry river beds, etc. that were easily recognizable on the ground51. Because of the ways in which temporary structures are erected throughout the communities, including on property belonging to permanent homes, an accurate enumeration of households was not possible. As such, before interviews commenced in each stratum, the enumeration maps were checked for accuracy using a handheld GPS and existing maps during a walkthrough of each stratum. This process checked for gross inaccuracies in terms of segment size. Maps were found to be relatively accurate. That is, while the population may have grown, the on the ground realities appeared to proportionally match the aerial maps.
Households were selected using a systematic sampling method accomplished via random walk. The research assistants started in the southwest corner of each stratum and surveyed every 5th household. Abandoned homes were not counted. Four households declined to participate, and in their place, the next household was surveyed. In total, one hundred households in each community were surveyed. Without knowing the exact population size, an a priori power analysis was run to determine the sample size needed for a small effect size, Cohen’s d = 0.4. A total sample size of n = 200 has enough power (80%) to detect this difference at the a = 0.05 probability level45.
Systematic sampling, a probability-based sampling method, was used because of its ease of application, especially in contexts where the sampling frame was unlikely to represent the communities as they currently existed, and where updating the sampling frame may be difficult, or nearly impossible, in some areas due to access, as it was in Stampriet and Gibeon. This sampling method, however, is associated with unequal probabilities of item, or household, inclusion. Nevertheless, systematic sampling behaves as simple random sampling and typically has the same precision for variables involving human populations 52.
All strata were sampled ensuring that the survey results were as representative of the entire community as possible. At the time of survey design and strata designation, the social structure and organization of each community was unknown to the researcher (e.g., stratification based on wealth, religion, or ethnicity), but it was assumed that stratification within the community exists, and that there is a high degree of homogeneity within each stratum 53. Thus, to yield a more precisely calculated sample mean for each variable, all residential strata were sampled.
Two Namibian research assistants were hired for the enactment of household interviews. A female interviewer was employed to maximize comfortability and ease of conversation with household respondents, the primary cooks, who were predominantly women or older children. A male driver was hired to assist with interviews, which were conducted primarily in Afrikaans, and to serve as an interpreter for interviews conducted in Khoekhoegwab, a local language in which he was fluent. Interviews were audio recorded, with participant permission, to allow for post-interview checking of respondent answers and for context. Recordings were later translated and transcribed.
Household Survey Specific Measures
The specific instruments included in the household questionnaire were chosen to make use of existing measures whenever possible. The aim of these instruments was to gather information about the primary cook’s knowledge, attitudes, and behaviours toward efficient cooking technology and sustainability practices within the home. In some cases, the instruments were not designed, nor necessarily intended, for non-Western contexts and are thus used in an exploratory way only. This is described further in subsequent sections below. Efforts were taken to ensure that key dependent variables are measured in numerous ways. For instance, participant responses about types of fuel use was asked in two different ways, both in terms of the frequency of use of each stove within the household as well as the number of meals prepared on the traditional stove each week. Questions on the youth survey were also used to confirm observations made in the communities.
Adoption Index Survey
Behaviours are both difficult to change and to quantify. To determine whether experience at NaDEET affected energy-related behavioural changes at the household level, the degree to which households adopt their electric and traditional cookstoves was examined, where adoption score was used as a proxy for behaviour. This survey is part of a toolkit developed for the Clean Cooking Alliance 43,54 and includes 8 questions regarding the user’s perceptions and reported use of an efficient cookstove, as well as a visual observation of the stoves used to confirm the participant’s responses. These questions were asked for each type of cookstove or fuel in the home and occurred throughout the interview, rather than as a discrete set of questions. Based on the responses, the adoption index was calculated for each stove in residence. The adoption index (AI) is calculated as a function of four variables: the frequency of use of the cookstove (FCCS), overall condition of the cookstove (CCCS), level of satisfaction with the cookstove (LSC), and her interest in replacing the cookstove with a similar one at the end of the cookstove’s lifetime (IRS).
Each variable mentioned above was given a score based on rubric that scores visual observations of the stove by the interviewer and the respondent’s answers to the questions. Based on a cluster analysis of several case studies, the variables were weighted and the following adoption equation for an individual stove was developed 43,54.
AI=4(FCCS)+3(CCCS)+2(LSC)+1(IRS) (1)
The index is meant to be flexible and allows for dropping of terms and alternative weights to provide a snapshot of stove uptake at a moment in time43. The condition term was dropped for electrical devices since tinkering or making modifications to an electrical device is outside the expertise of most users. This term, as well as the interest in replacing term are irrelevant in the context of traditional stoves, and thus also dropped from the equation.
Electric stove commitment was thus measured by equation 4
4(FCCS) + 2(LSC) + (IRS) (4)
and traditional stove commitment by equation 5.
4(FCCS) + 2(LSC) (5)
Adoption scores ranged from 0 to 7 for electric stoves, and up to 6 for traditional cookstoves (Table 4). A one-point difference between scores, based on the variables and weightings used, represents an additional two or three days of cookstove use per week, or an increase from ambivalence to high satisfaction with the stove’s performance. To compare electric and traditional stove commitments, the index scores were normalized such that commitment was viewed as the proportion of full adoption according to the values listed in Table 5.
Attitudes about Solar Cooking
Mercy et al. (2008)55 developed a short questionnaire for assessing women’s perceptions and knowledge about solar cookers in Mali. These 10 questions are based on a 5-point Likert scale. Minor adjustments were made to the survey to replace references to Mali, the location of the instrument development, with Namibia.
Six Americas Short Survey (SASSY)
The Six America’s Global Warming survey consists of 36 questions which assess a respondent’s knowledge and attitudes about global warming and climate change. This survey has been used since 2008 to segment the American population into six groups based on their beliefs, attitudes, and level of concern about global warming 56. The results of this instrument have been used in a variety of ways by researchers, educators, and policy makers. Most recently there has been interest in its use for tailoring communication about climate change to specific audiences 56. Several other countries have used or adapted this instrument, including a British Broadcasting Corportation survey of 33,000 residents from six countries in Asia with the explicit goal of improving their communication strategies 57. The instrument has also been used on small, sub-groups such as farmers in the corn belt of the United States 58, but to date, no studies exist in which rural residents of LMICs have been surveyed. A subset of four questions has been used to reliably assess an individual’s perceptions about global warming risks, expected harm to future generations, and how important the respondent finds these issues, as accurately as if the entire instrument was used 56. These four questions were asked at the very end of the questionnaire as to not introduce bias into solar energy responses earlier in the interview.
Sampling Weights
Sampling weights were calculated to account for differential probabilities of selection based on unequal strata sizes in terms of number of households per stratum, to improve precision of mean population estimations. The sampling frame and data from the most recent Namibian census was used to determine the population size in individual, or groups, of strata used in this study to calculate the weights. For instance, according to the census, there were 513 people living in strata 1, 2, 3, and 4 in Gibeon in 2011. This data, combined with household size data collected during the study, was used to calculate sampling weights for each stratum. The total number of people in the houses sampled in these four areas was n = 196, yielding a base weight, the inverse of probability of selection, wi = 2.62.
There were 4 nonresponses; 1 in Stampriet, and 3 in two different segments in Gibeon. Nonresponse weights,
where Ss is the number of cases sampled for the segment and Sp is the number of responses obtained for the segment, were calculated for segments containing a nonresponse. Nonresponse weights were multiplied by the base weight for an adjusted base weight. The relative weights, were then found by dividing the adjusted base weights by 
the mean of all adjusted base weights for the community,
to yield the total sampling weight53. The survey design was declared in Stata 16 using these total sampling weights which were then used for all population mean estimations using Stata’s svy commands for more precise estimate means and confidence intervals.
All reported independent samples and paired t-tests, and two proportion z-tests are two-tailed.
NaDEET Learner Survey
Pre-Survey
Two weeks before the students’ visit to NaDEET Centre, surveys as well as consent documents and teacher instructions were sent to participating schools via NamPost Courier, the courier arm of national postal service. Teachers were instructed to read the questionnaire out loud to students as a group in English, the language of instruction throughout Namibia, or in the children’s primary language, as needed. Students recorded their answers on the questionnaire provided. Teachers were asked to check the questionnaires for completeness as the students turned them in to minimize missing data. The pre-survey consisted of demographic items and questions to establish the students’ baseline knowledge and attitudes about residential energy and other sustainability-related concerns. The same solar energy questions asked of adults in the community survey were included on the student questionnaire (Supplementary Note 3). Like the community survey, this questionnaire used a mix of existing instruments and original questions written in collaboration with NaDEET leadership. Teachers were provided with a pre-paid envelope to return the completed surveys via courier.
Post-Survey and Six-Month Follow-Up
Students completed a post-survey two weeks after their visit to NaDEET Centre. The post-survey questionnaire was a shorter version of the pre-survey. Teachers returned the post-survey in the same pre-paid courier parcel as the pre-survey. Follow-up surveys were sent to schools approximately six months after their visit to NaDEET, again using NamPost Courier.
Control Group
Each participating school was asked to identify a class of students one grade below the students scheduled to attend NaDEET to serve as a control group. This ensured that the control group was similar to the NaDEET participants demographically, eventually eligible for the same opportunities, but without prior NaDEET experience. Surveys were administered to both NaDEET learners and the control group on the same schedule.
Missing Data
Missing data was handled in Stata using valid mean substitution (VMS)59, a method appropriate for Likert-scale attitude items in which the minimum and maximum of all items are the same, and thus the theoretical means and standard deviations are equal60. When data is found to be missing at random, as it was in this study, VMS has been found to produce similar estimates as multiple imputation methods61.
Both studies were approved by the [blinded university]’s Institutional Review Board. The Namibia Commission for Research, Science, and Technology and the Ministry of Education also issued approvals for these studies.
Limitations
While every effort was taken to systematically sample the entire town to achieve a representative sample of each community, it is possible that groups of households were missed due to the random walk method employed during sampling, interviewer errors, or households that were inaccessible or unobservable from the primary residential areas. This may be especially true for informal settlements outside of the neighbourhood centres. Two Namibian research assistants who speak fluent Afrikaans were employed to conduct the household interviews to minimize social desirability bias62–64. Given that the research assistants were in each community for approximately two weeks, it is likely that some respondents, especially those interviewed later in the process, already knew who the research assistants were and which organizations they represented.
Human behaviours and attitudes are informed by a tapestry of interwoven variables and influences. While there is evidence that points to NaDEET’s impact at the household and community level, there is no way to know with certainty if observed differences are due to learning accrued at NaDEET, or if there is some other factor that was not uncovered by the interviews and time spent in these communities. Moreover, the researcher and her assistants were outsiders to the communities. Some nuance in responses may have not been detected due to a lack of understanding of cultural context, shared experience, and interpretation.
Troncoso’s adoption index was used because it was already developed and promoted by the Clean Cooking Alliance, but there are some issues with it as an instrument, and though its developers advocate for its flexibility in use, it was heavily adapted for this study. Adoption as a construct is complex and this formula considers just four factors. For instance, while the index intimates fuel stacking by incorporating the frequency of use of the cookstove in question, it does not consider the degree of fuel stacking within the household, nor a stove’s importance relative to other devices in the home. The instrument itself does not capture what the cookstove is replacing, or how many other fuels and stoves are used. In an attempt to address these limitations, the index was used for each stove in the household, but more work is needed in this area.
The long-distance nature of the youth survey adds to the limitations that exist by virtue of the methods employed. The reliance on a national courier system to deliver surveys across the vast Namibian landscape was more successful than expected. However, a few schools did not receive the surveys in the intended time frame, or at all, which made acquiring three sets of data points from each student impossible even if the COVID-19 global pandemic hadn’t closed schools in early 2020, disrupting the third round of data collection.
- Namibia Statistics Agency. Namibia 2011 Population and Housing Census Main Report. http://www.nsa.org.na/files/downloads/Namibia 2011 Population and Housing Census Main Report.pdf (2011).
- Global Alliance for Clean Cookstoves. Country Profile: Namibia. http://cleancookstoves.org/country-profiles/29-namibia.html (2017).
- GBD 2016 Risk Factors Collaborators. Global, regional, and national comparative risk assessment of 84 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990 – 206: a systematic analysis for the Global Burden of Disease Study 2016. Lancet 390, 1345–1422 (2017).
- International Energy Agency, International Rewnewable Energy Agency, United Nations Statistics Division, World Bank & World Health Organization. Tracking SDG7: The Energy Progress Report 2020. (2020).
- GBD 2019 Risk Factors Collaborators. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 396, 1233–1249 (2020).
- Sovacool, B. K. What are we doing here? Analyzing fifteen years of energy scholarship and proposing a social science research agenda. Energy Res. Soc. Sci. 1, 1–29 (2014).
- Reyes, R., Schueftan, A., Ruiz, C. & González, A. Controlling air pollution in a context of high energy poverty levels in southern Chile: Clean air but colder houses? Energy Policy 124, 301–311 (2019).
- Samuel, O., Baba, M. & Thlimabari, A. The Ecological, Economical and Social Impact of Adoption Solar Cooking in Mubi Metropolis Adamawa State, Nigeria. J. Sci. Res. Reports 1–6 (2018).
- Bond, T. C. et al. Bounding the role of black carbon in the climate system: A scientific assessment. J. Geophys. Res. Atmos. 118, 5380–5552 (2013).
- Palmer, C. & MacGregor, J. Fuelwood Scarcity, Energy Substitution and Rural Livelihoods in Namibia. Environ. Dev. Econ. 14, 693–715 (2009).
- Gebreegziabher, Z., van Kooten, G. C. & van Soest, D. P. Technological innovation and dispersion: Environmental benefits and the adoption of improved biomass cookstoves in Tigrai, northern Ethiopia. Energy Econ. 67, 337–345 (2017).
- Masera, O. R., Díaz, R. & Berrueta, V. From cookstoves to cooking systems: the integrated program on sustainable household energy use in Mexico. Energy Sustain. Dev. 9, 25–36 (2005).
- Bazilian, M., Cordes, L., Nussbaumer, P. & Yager, A. Partnerships for access to modern cooking fuels and technologies. Curr. Opin. Environ. Sustain. 3, 254–259 (2011).
- Ezzati, M. et al. Energy Management and Global Health. Annu. Rev. Environ. Resour. 29, 383–419 (2004).
- Shankar, A. et al. Maximizing the benefits of improved cookstoves: Moving from acquisition to correct and consistent use. Global Health Science and Practice vol. 2 268–274 (2014).
- Rosa, G. et al. Assessing the impact of water filters and improved cook stoves on drinking water quality and household air pollution: A randomised controlled trial in Rwanda. PLoS One 9, e91011 (2014).
- Miller, G. & Mushfiq Mobarak, A. Learning about new technologies through social networks: Experimental evidence on nontraditional stoves in Bangladesh. Mark. Sci. 34, 480–489 (2015).
- Iessa, L., De Vries, Y. A., Swinkels, C. E., Smits, M. & Butijn, C. A. A. What’s cooking? Unverified assumptions, overlooking of local needs and pro-solution biases in the solar cooking literature. Energy Res. Soc. Sci. 28, 98–108 (2017).
- Jeuland, M. A. & Pattanayak, S. K. Benefits and costs of improved cookstoves: Assessing the implications of variability in health, forest and climate impacts. PLoS One 7, e30338 (2012).
- Malla, S. & Timilsina, G. Household cooking fuel choice and adoption of improved cookstoves in developing countries: a review. (2014).
- Mobarak, A. A. M., Dwivedi, P., Bailis, R., Hildemann, L. & Miller, G. Low demand for nontraditional cookstove technologies. Proc. Natl. Acad. Sci. 109, 10815–10820 (2012).
- Rehfuess, E. A., Puzzolo, E., Stanistreet, D., Pope, D. & Bruce, N. G. Enablers and barriers to large-scale uptake of improved solid fuel stoves: A systematic review. Environ. Health Perspect. 122, 120–130 (2014).
- Abdelnour, S., Pemberton-Pigott, C. & Deichmann, D. Clean cooking interventions: Towards user-centered contexts of use design. Energy Res. Soc. Sci. 70, 101758 (2020).
- Gill-Wiehl, A., Price, T. & Kammen, D. What’s in a stove? A review of the user preferences in improved stove designs. Energy Res. Soc. Sci. 81, 102281 (2021).
- Lindgren, S. Clean cooking for all? A critical review of behavior, stakeholder engagement, and adoption for the global diffusion of improved cookstoves. Energy Res. Soc. Sci. 68, 101539 (2020).
- Boudet, H. et al. Effects of a behaviour change intervention for Girl Scouts on child and parent energy-saving behaviours. Nat. Energy 1, 1–10 (2016).
- Kleinschafer, J. & Morrison, M. Household norms and their role in reducing household electricity consumption. Int. J. Consum. Stud. 38, 75–81 (2014).
- Kar, A. & Zerriffi, H. From cookstove acquisition to cooking transition: Framing the behavioural aspects of cookstove interventions. Energy Res. Soc. Sci. 42, 23–33 (2018).
- Pilloni, M., Hamed, T. & Joyce, S. Assessing the success and failure of biogas units in Israel: Social niches, practices, and transitions among Bedouin villages. Energy Res. Soc. Sci. 61, 101328 (2020).
- Ochieng, C., Murray, U., Owuor, J. & Spillane, C. The forgotten half: Men’s influence over cookstove adoption decisions in Northern Kenya. Energy Res. Soc. Sci. 74, 101913 (2021).
- Bandura, A. Social Foundations of Thought and Action: A Social Cognitive Theory. The National Institute of Mental Health (1986). doi:10.4135/9781446221129.n6.
- Okyere, C. Y., Pangaribowo, E. H., Asante, F. A. & von Braun, J. The Impacts of Household Water Quality Testing and Information on Safe Water Behaviors: Evidence from a Randomized Experiment in Ghana. SSRN Electron. J. (2017) doi:10.2139/ssrn.2955038.
- Maio, G. R. et al. Social Psychological Factors in Lifestyle Change and Their Relevance to Policy. Soc. Issues Policy Rev. 1, 99–137 (2007).
- Bresee, S., Caruso, B., Sales, J., Lupele, J. & Freeman, M. ‘A child is also a teacher’: exploring the potential for children as change agents in the context of a school-based WASH intervention in rural Eastern Zambia. Health Educ. Res. 31, 521–534 (2016).
- Kepha, S. et al. Plasmodium falciparum parasitaemia and clinical malaria among school children living in a high transmission setting in western Kenya. Malar. J. 15, 157 (2016).
- Mahanta, T. et al. Effect of social and behavior change communication by using infotainment in community perception of adolescent girls for reproductive and sexual health care in high priority districts of Assam. Clin. Epidemiol. Glob. Heal. 4, 133–139 (2016).
- Puffer, E. et al. A church-based intervention for families to promote mental health and prevent HIV among adolescents in rural Kenya: Results of a randomized trial. J. Consult. Clin. psych2ology 84, 511 (2016).
- Kandpal, T. C. & Broman, L. Renewable energy education: A global status review. Renew. Sustain. Energy Rev. 34, 300–324 (2014).
- Lane, J. F., Floress, K. & Rickert, M. Development of school energy policy and energy education plans: A comparative case study in three wisconsin school communities. Energy Policy 65, 323–331 (2014).
- Ntona, E., Arabatzis, G. & Kyriakopoulos, G. L. Energy saving: Views and attitudes of students in secondary education. Renew. Sustain. Energy Rev. 46, 1–15 (2015).
- Damerell, P., Howe, C. & Milner-Gulland, E. J. Child-orientated environmental education influences adult knowledge and household behaviour. Environ. Res. Lett. 8, (2013).
- Shankar, A. V. et al. Everybody stacks: Lessons from household energy case studies to inform design principles for clean energy transitions. Energy Policy 141, 111468 (2020).
- Troncoso, K. A Recipe for Developing Adoption & Impact Indices. (2013).
- Guta, D. et al. A systematic review of household energy transition in low and middle income countries. Energy Res. Soc. Sci. 86, 102463 (2022).
- Cohen J. Statistical Power Analysis for the Behavioural Science (2nd Edition). Statistical Power Anaylsis for the Behavioral Sciences (1988).
- Rehfuess, E. A., Puzzolo, E., Stanistreet, D., Pope, D. & Bruce, N. G. Enablers and barriers to large-scale uptake of improved solid fuel stoves: A systematic review. Environ. Health Perspect. 122, 120–130 (2014).
- Diaz-Maurin, F., Chiguvare, Z. & Gope, G. Scarcity in abundance: The challenges of promoting energy access in the Southern African region. Energy Policy 120, 110–120 (2018).
- Alweendo, T. Press statement on the power supply situtation in Namibia. (2018).
- Cooper, I. D. & Johnson, T. P. How to use survey results. Journal of the Medical Library Association (2016) doi:10.3163/1536-5050.104.2.016.
- United Nations. Transforming Our World: The 2030 Agenda for Sustainable Development. A New Era in Global Health (2015) doi:10.1891/9780826190123.ap02.
- Eckman, S., K, H. & Dever, J. Innovative Sample Designs Using GIS Technology. in Advances in Comparative Survey Methods: Multinational, Multiregional, and Multicultural Contexts (3MC) 67 (2018).
- Sudman, S. Applied Sampling. (Academic Press, Inc., 1976).
- Lohr, S. Sampling: design and analysis. (Nelson Education, 1999).
- Troncoso, K., Armendáriz, C. & Alatorre, S. Improved cook stove adoption and impact assessment: A proposed methodology. Energy Policy 62, 637–645 (2013).
- Mercy, A., Asinobi, C. & Yemi, I. Attitudes of Rural and Urban Women in Bamako District and Koulikoro Region of Mali to the Use of a Solar Cooker. in 24th Annual Meeting of the Association for International Agricultural and Extension Education 15–27 (2008).
- Chryst, B. et al. Global Warming’s “Six Americas Short Survey”: Audience Segmentation of Climate Change Views Using a Four Question Instrument. Environ. Commun. (2018) doi:10.1080/17524032.2018.1508047.
- BBC Media Action. Climate Asia methodology series. Regional segment profiles. (2013).
- Arbuckle, J. G. et al. Climate change beliefs, concerns, and attitudes toward adaptation and mitigation among farmers in the Midwestern United States. Clim. Change (2013) doi:10.1007/s10584-013-0707-6.
- Raymond, M. R. Missing data in evaluation research. Eval. Health Prof. 9, 395–420 (1986).
- Dodeen, H. M. Effectiveness of valid mean subsitution in treating missing data in attitude assessment. Assess. Eval. High. Educ. 28, 505–513 (2003).
- Raaijmakers, Q. Effectiveness of different missing data treatments with Likert-type data: Introducing the relative mean substitution approach. Educ. Psychol. Meas. 59, 725–748 (1999).
- Nederhof, A. J. Methods of coping with social desirability bias: A review. Eur. J. Soc. Psychol. (1985) doi:10.1002/ejsp.2420150303.
- Moses, N. & MacCarty, N. Cookstove Usability Testing Protocol. (2018).
- Weiss, R. S. Learning From Strangers. Qual Interviews.Pdf. Learning from strangers The art and method of qualitative interview studies (1994).
|
|
Stampriet Mean (sd) |
Gibeon Mean (sd) |
z |
df |
p |
|
Access to electrical grid |
0.97 (0.22) |
0.84 (0.46) |
2.65 |
|
.008 |
|
Electric stove ownership |
0.72 (0.69) |
0.86 (0.48) |
1.70 |
|
.088 |
|
Daily electric stove use |
0.55 (0.54) |
0.83 (0.45) |
4.02 |
|
.001* |
|
Monthly electricity expenditures ($N) |
275.34 (290.29) |
391.50 (314.66) |
2.71† |
173 |
.007 |
Table 1. Estimated electricity usage in Stampriet and Gibeon. Two proportion z-tests results to compare electricity access and usage in Stampriet and Gibeon for households connected to the municipal grid.
* Indicates p-values less than 0.001
† t-test scores as a result of a two-tailed independent samples t-test.
|
|
|
B |
SE |
b |
t |
p |
||
|
Electric |
Socioeconomic Status |
0.43 |
0.13 |
0.29 |
3.31 |
.001* |
||
|
|
Level of Education |
0.01 |
0.05 |
0.01 |
0.13 |
.897 |
||
|
|
NaDEET Experience |
-0.25 |
0.37 |
-0.06 |
-0.67 |
.502 |
||
|
|
F(3, 135) = 4.49, p = .005, R2 = 0.09
|
|
|
|
||||
|
Traditional
|
Socioeconomic Status Level of Education NaDEET Experience |
-0.38 -0.10 -0.72 |
0.08 0.04 0.31 |
-0.32 -0.18 -0.15 |
-4.44 -2.48 -2.34 |
.001* .014 .020 |
||
|
|
F(3, 188) = 17.61, p < .001, R2 = 0.22
|
|
|
|
|
|||
Table 2. Summary of regression analysis for variables predicting electric and traditional cookstove commitment.
* Indicates p-values less than .001
|
|
NaDEET Mean (sd) |
Control Mean (sd) |
t(387) |
p |
|
Age |
13.04 (1.13) |
12.44 (1.08) |
4.91 |
.001* |
|
Female Student |
0.61 (0.49) |
0.52 (0.50) |
0.76† |
.446 |
|
Family Size |
6.60 (2.71) |
6.85 (2.77) |
0.79 |
.428 |
|
Electricity at Home |
0.91 (0.28) |
0.97 (0.18) |
0.48† |
.631 |
|
Prior Household NaDEET Experience |
0.63 (0.48) |
0.71 (0.46) |
0.70† |
.483 |
Table 3. Overview of NaDEET participants and students in the control group across key demographical variables. Two schools attended NaDEET Centre with little advance notice and pre-survey demographic information is missing from those students, and therefore not included in this table.
* Indicates p-values less than .001
† z-test scores from two-proportion tests of binary variables
|
|
Value |
||||
|
0 |
0.25 |
0.5 |
0.75 |
1.0 |
|
|
Frequency of use (FCCS) |
Never |
Once per week or less |
2 to 3 days per week |
4 to 6 days per week |
Every day |
|
Condition (CCCS) |
Destroyed or in disuse |
Modifications; performance impairment |
Modifications; no performance impairment |
Working with low maintenance |
Perfect with good maintenance |
|
Level of satisfaction (LSC) |
Unsatisfied |
Low satisfaction |
Regularly satisfied |
Satisfied |
Very satisfied |
|
Interest in replacing (IRS) |
No |
|
Maybe |
|
Yes |
Table 4. Troncoso's rubric for scoring adoption index variables43.
|
Percentage of Total Points |
Adoption Description |
|
0.90 – 1.0 |
Very Good Adoption |
|
0.85 > 0.9 |
Good Adoption |
|
0.70 > 0.85 |
Regular/Moderate Adoption |
|
0.50 > 0.70 |
Bad Adoption |
|
0.0 > 0.50 |
Very Bad Adoption |
Table 5. Modified Adoption Index scoring guide used for electric and traditional cookstoves in this study43. The total points available are determined from Equations 4 and 5 and the rubric shown by Table 4.
There is NO Competing Interest.