Dear forum members,
We have a panel data set of 15-min interval electric vehicle charging data. Since we do not observe household electricity consumption, we have lots of true zeroes (where people don't charge their car).
Furthermore, we have treatments (T1, T2, T3) which are different tariffs that people had for 3 months each. In each tariff, people had different charging conditions (higher/lower prices, lower charging speed, etc.) for certain time periods during some days of the respective treatment period.
The tariffs were given to the respondents in different orders, meaning that one group got T1, then T3, then T2 and another group got T2, then T1, then T3, etc. T1 is supposed to act as our control group.
Our dependent variable will be the energy charged per 15-min interval, and what we want to conclude whether people shift their charging in time due to the different charging conditions in the contracts (e.g., whether people charge less when the price is higher at 8pm during some days).
We are unsure what the best way to analyze this data would be, especially due to the shifted treatment times and the large number of true zeroes.
I would be very grateful for any suggestions or tips on how to work with this sort of data.
Kind regards
We have a panel data set of 15-min interval electric vehicle charging data. Since we do not observe household electricity consumption, we have lots of true zeroes (where people don't charge their car).
Furthermore, we have treatments (T1, T2, T3) which are different tariffs that people had for 3 months each. In each tariff, people had different charging conditions (higher/lower prices, lower charging speed, etc.) for certain time periods during some days of the respective treatment period.
The tariffs were given to the respondents in different orders, meaning that one group got T1, then T3, then T2 and another group got T2, then T1, then T3, etc. T1 is supposed to act as our control group.
Our dependent variable will be the energy charged per 15-min interval, and what we want to conclude whether people shift their charging in time due to the different charging conditions in the contracts (e.g., whether people charge less when the price is higher at 8pm during some days).
We are unsure what the best way to analyze this data would be, especially due to the shifted treatment times and the large number of true zeroes.
I would be very grateful for any suggestions or tips on how to work with this sort of data.
Kind regards
