Ive started to design the survey with demographics, I have a list of pro-environmental and conservation behaviours which i plan to ask if they have changed since the pandemic. I think i'll use likert scale on those questions eg: Since the pandemic I have recycled more. Strongly agree, agree etc. I plan to then have a couple of quantitative questions at the end about the reasons for the change in their behaviours. What would be the best statistical analysis for this? Also, I'm confused with the hypothesis and variable. Could their employment status be an independent variable? or should i just be focussing on the pandemic being the independent variable and behaviour being the dependent variable. Thank you
Dear Sally Freemantle,
In my opinion, in the survey conducted on the issue of the impact of the SARS-CoV-2 (Covid-19) coronavirus pandemic on the change of pro-environmental behavior of people, the key issue is to examine the potential increase in the general pro-environmental awareness of people and what issues linking the SARS-CoV- coronavirus pandemic 2 (Covid-19) with the issue of the need to improve the protection of the natural environment, protect the biodiversity of natural ecosystems, protect the biosphere and climate of the planet Earth, people indicate as the main factors connecting these two issues. Various studies currently conducted indicate that there are numerous connections between the two. On the other hand, in individual communities, these may be different connecting factors present in the general social consciousness.
Best regards,
Dariusz Prokopowicz
You cannot treat the pandemic as an independent variable in a statistical analysis because you don't plan to measure it.
Overall, it sounds like you need to consult an introductory text on surveys, such as Fowler, Survey Research Methods.
Dear Sally Freemantle ,
I think your null hypothesis is that views have not changed since the pandemic. This is slightly unusual as normally you would compare two groups or the relationship between two variables. You also need to consider whether you wish to analyse questions individually or collectively. If the latter then this may lead your in a dimension reduction technique such as a Principal Component Analysis with a scale reliability analysis. If you were to score your central value of your scale as zero then you could carry out a t-test paired with 0.
Another alternative would be to seek to order your individual question responses using a Friedman's ANOVA.
- Badges
- Science topic