Academia.eduAcademia.edu

Figure 3 – uploaded by Theresa Tawiah

See full PDFdownloadDownload figure
Figure 2: Box plots for wealth score distribution for the three transition choices. Respondents with successful outcomes are indicated with green dots, while those with failure outcomes are red for each choice node. For clarity on the difference across the processes in the plot, we have only displayed the wealth index score up to 14 (max. score 48.705). Effect size (Hedge’s d) is also shown. The shade of dots for uptake looks brighter due to the intensity- more respondents (points) per unit area.

Figure 2 Box plots for wealth score distribution for the three transition choices. Respondents with successful outcomes are indicated with green dots, while those with failure outcomes are red for each choice node. For clarity on the difference across the processes in the plot, we have only displayed the wealth index score up to 14 (max. score 48.705). Effect size (Hedge’s d) is also shown. The shade of dots for uptake looks brighter due to the intensity- more respondents (points) per unit area.

Related Figures (7)

Figure 1: Conceptual framework of four stages (circle) of the clean cooking transition with the three transition choices (boxes): The three transition choices: LPG uptake (1), LPG as main fuel (2), and LPG as exclusive fuel (3) decide the outcome: the present stage of a household out of the four options. While this is here drawn as sequential and unidirectional for simplicity, we expect both stage-skipping as well as retrograde motion in some cases. Given the aforementioned gaps in our knowledge, we propose and test a novel conceptual framework of transition. We suggest that there are four distinct stages in the transition process from solid fuels to LPG over time (Figure 1): households typically begin their journey at stage (1) no LPG use, i.e. exclusive use of solid fuels. Then some households progress to stage (2) where they use LPG as a non-primary (secondary or tertiary) fuel while continuing to use solid fuels as primary cooking fuel. Few households then progress to the stage (3) where they start the use of LPG as the primary cooking fuel. Finally, households that completely disadopt solid fuels reach the stage (4) when they exclusively use LPG. We posit that the factors that predict the type of fuel households use, as well as their level of influence in such prediction, change depending on which stage the household is in the clean cooking energy transition journey.
Figure 1: Conceptual framework of four stages (circle) of the clean cooking transition with the three transition choices (boxes): The three transition choices: LPG uptake (1), LPG as main fuel (2), and LPG as exclusive fuel (3) decide the outcome: the present stage of a household out of the four options. While this is here drawn as sequential and unidirectional for simplicity, we expect both stage-skipping as well as retrograde motion in some cases. Given the aforementioned gaps in our knowledge, we propose and test a novel conceptual framework of transition. We suggest that there are four distinct stages in the transition process from solid fuels to LPG over time (Figure 1): households typically begin their journey at stage (1) no LPG use, i.e. exclusive use of solid fuels. Then some households progress to stage (2) where they use LPG as a non-primary (secondary or tertiary) fuel while continuing to use solid fuels as primary cooking fuel. Few households then progress to the stage (3) where they start the use of LPG as the primary cooking fuel. Finally, households that completely disadopt solid fuels reach the stage (4) when they exclusively use LPG. We posit that the factors that predict the type of fuel households use, as well as their level of influence in such prediction, change depending on which stage the household is in the clean cooking energy transition journey.
Table 1: Thirteen factors identified as drivers and barriers based on a review of published literature on the clean cooking energy transition factors
Table 1: Thirteen factors identified as drivers and barriers based on a review of published literature on the clean cooking energy transition factors
2. Key predictors are unlikely to drive success across all transition choices
2. Key predictors are unlikely to drive success across all transition choices
Figure 3: Stacked barplot to describe the relationship between urbanicity and LPG transition for the three transition choices: uptake, main and exclusive. The percentage of respondents who have success (green) and failure (red) are grouped by urbanicity (urban/ rural) along with the Pearson Chi-Square test results. Cramér's V denoting the strength of association is also plotted.
Figure 3: Stacked barplot to describe the relationship between urbanicity and LPG transition for the three transition choices: uptake, main and exclusive. The percentage of respondents who have success (green) and failure (red) are grouped by urbanicity (urban/ rural) along with the Pearson Chi-Square test results. Cramér's V denoting the strength of association is also plotted.
Figure 4: Coefficient plot for all the three processes (ref: Figure 1) of LPG transition only showing predictors that were found to be statistically significant for the three choices. exp(Estimate) provides the odds value of the factor using the exponential of the estimate (coefficient value) of the binary logistic regression. The circles represent the odds value while the line shows the 95% confidence interval around the odds value. The coefficient plot for binary logistic regression for the three processes is plottec in Figure 4 where only the influential [statistically significant, p<0.05] factors are shown. The figure shows the odds, with exp(Estimate) equal to the exponential of coefficient estimate (log odds) from the regression summary. The full regression results for all the thirteen factors are available in Supplementary Figure S111.
Figure 4: Coefficient plot for all the three processes (ref: Figure 1) of LPG transition only showing predictors that were found to be statistically significant for the three choices. exp(Estimate) provides the odds value of the factor using the exponential of the estimate (coefficient value) of the binary logistic regression. The circles represent the odds value while the line shows the 95% confidence interval around the odds value. The coefficient plot for binary logistic regression for the three processes is plottec in Figure 4 where only the influential [statistically significant, p<0.05] factors are shown. The figure shows the odds, with exp(Estimate) equal to the exponential of coefficient estimate (log odds) from the regression summary. The full regression results for all the thirteen factors are available in Supplementary Figure S111.
Figure 5: Qualitative (self-assessed LPG use level) response vs. Quantitative (self-reported) responses. The median value (m) and sample size (n) are shown for each indicator for each of the three levels for LPG users. While m and n values are calculated for all respondents, for plotting boxplots we have used a modified dataset. We removed outliers (>99 percentile) and then generated boxplots to avoid visual cluttering. parameters on LPG use. We mapped the qualitative response (level of LPG use) of LPG users with five different quantitative findings with varying parameters and timescales: use [seasonal and yesterday] use (min/day), [weekly] frequency of use (times/ week), [last year] LPG purchase (kg/ year), and [last month] LPG purchase (Ghana Cedi/ month). At the individual household level, these responses, even if carefully answered, would be subject to recall bias and depend on the respondent’s sense of time. Hence, we examined the consistency of the overall survey findings instead of assessing accuracy at the household level. We note that there is overall consistency between quantitative measures of LPG use and the qualitative assessmen (Figure 5). This provides reasonable assurance on the overall consistency of respondents’ self-assessments of the stage of transition.
Figure 5: Qualitative (self-assessed LPG use level) response vs. Quantitative (self-reported) responses. The median value (m) and sample size (n) are shown for each indicator for each of the three levels for LPG users. While m and n values are calculated for all respondents, for plotting boxplots we have used a modified dataset. We removed outliers (>99 percentile) and then generated boxplots to avoid visual cluttering. parameters on LPG use. We mapped the qualitative response (level of LPG use) of LPG users with five different quantitative findings with varying parameters and timescales: use [seasonal and yesterday] use (min/day), [weekly] frequency of use (times/ week), [last year] LPG purchase (kg/ year), and [last month] LPG purchase (Ghana Cedi/ month). At the individual household level, these responses, even if carefully answered, would be subject to recall bias and depend on the respondent’s sense of time. Hence, we examined the consistency of the overall survey findings instead of assessing accuracy at the household level. We note that there is overall consistency between quantitative measures of LPG use and the qualitative assessmen (Figure 5). This provides reasonable assurance on the overall consistency of respondents’ self-assessments of the stage of transition.
Related topics:
Environmental HealthPsychological InterventionLiquefied petroleum gas
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved