We are making progress! This week we scripted two surveys, one for employers and one for employees. Because the surveys will be orally administered, we are making sure not to ask any personal details. We worked to ensure that the questions are concise and scripted dialogue at the beginning of each survey informing about voice-recording permission and participants’ right to decline any answer. Most of the questions in the survey are simple and can be answered with a quick one or two word response. For questions that require a numerical answer or questions that necessarily impose choosing from a scale, we have scripted broad ranges to help participants. A few of the questions involve thinking back to commute patterns and times and we have discussed at length how best to phrase these questions as simple as possible while still making them sufficiently detailed. We aimed to keep the surveys around ten to fifteen minutes in length. However, since we haven’t administered a survey, we don’t yet know how long they might take, neither do we know how long of a survey an employer or employee is willing to go through. Wendy and Adina suggested slight modifications to the surveys. The most crucial change is to modify a few questions regarding how much the employers would be willing to pay for the transit subsidy after the TMA pilot program ends. We will implement these changes in the next day and will finalize all Spanish translations as well.
In order to respect the privacy of participants while still garnering useful data for PATMA, we have written a separate survey questionnaire in which both employers and employees may enter sensitive demographic data. The idea of separating the sensitive and general information into an oral and written portions comes from the Wednesday’s guest speaker, Vivian Chang. The questionnaires will be identified with a random ID number so if one ever gets misplaced, its content cannot be traced to a specific person. We will store a table that matches the questionnaires to the oral surveys in a secure, offline location to prevent accidental sharing on Google Drive.
We have begun preliminary comparison of the Redwood City Employee Survey and the Palo Alto Downtown Mode Split survey. This Redwood City data is the most comprehensive study out of all the studies that Jessica Manzi provided us. We now have a broad notion of what exactly one dataset surveyed that the other did not. From here we can springboard into figuring out which metrics Redwood City must address in future surveys.
What We Observed and Learned
This week, our meeting with Adina was extremely informative and helpful for our next steps, and also addressed and assuaged some concerns we had with the survey. We were struggling with how to ask questions about the hypothetical subsidized passes in a way that was both easy to understand and would provide us with the sort of answers we needed. We had settled on naming a cost in the middle of the range spanned by the various transit carriers around the Bay, but of course that is not realistic for everyone since the carriers people use and the cost of passes will vary depending on where people commute from. Adina helped us solve that problem by suggesting we use Google or Apple Maps to investigate people’s commute options on the spot. If a person gives us an intersection or landmark near their home, and we know the time they need to arrive to work, we can easily show them what transit carrier they could use, how long it would take compared to driving, and how much it would cost per trip. This would be pretty simple to implement during the survey, since we will have at least two people interviewing at a time. While one person covers the earlier part of the survey, another person can search what transit agencies the employee would use, since transit agencies have different pricing structures.
This feeds into another discovery: the difference between qualitative and quantitative surveys. Since these surveys are very qualitative, customizing questions for each respondent and having more of a conversation is fine, as long as we are able to get at the information we need one way or another. Adina told us we should expect to tweak the survey after administering it once or twice, and that making changes at that point is not only okay but best practice. This is very different from a more quantitative survey, where the questions must be standardized and no deviation from the script is allowed because it would affect the data. As we move into actual surveying, it will be important to be mindful the qualitative nature of the survey and not get bogged down in trying to get “clean” responses like we’d need for a quantitative study.
Adina also helped with how to pick businesses to survey. She suggested choosing those that Wendy suggested (which we have done) in addition to picking other business more-or-less randomly. She also clarified that two weeks ago when she said we should aim for 6-20 surveys, she meant surveying 6-20 businesses. While the lower end of this range shouldn’t be a problem, 20 business each with two to five employees might be a stretch since we are time limited. Once we start administering surveys, we will get a better sense of how many businesses we can survey, and adapt our goals based on that. We do have a couple survey meetings set up already though; we are surveying the manager of the Philz Coffee on Forest on Tuesday, and are in the process of arranging a survey of one of his employees as well. Our group has also determined time slots when our team can conduct fieldwork in a pair or trio. In order to make the best use of our time, we will try to conduct fieldwork for at least two or three hours at a time, since it does take a while to back and forth between campus and downtown Palo Alto.
Finally, as we’ve begun our analysis of Redwood City’s data, we’ve found that comparing it directly to Palo Alto’s is like fitting a round peg in a square hole. The surveys get at many of the same questions, but do it in very different ways, so it’s difficult to draw direct comparisons. Also, Redwood City has five different surveys to Palo Alto’s one data set. We may have to rethink our strategy for comparing the datasets, and approach it less by looking at what variables each survey covers and more by what questions, broadly, each survey addresses.
Critical Analysis / Moving Forward
We have a lot of momentum going forward on the Palo Alto project. This is good, since we still have a lot of work to do, and we have more clarity on how to approach the project. There are still some key changes that we will make to our surveys’ content. First, we still need to translate the employer survey to Spanish. We had assumed that business owners would most likely speak English; however, we realized that some owners may be more comfortable speaking in Spanish rather than English, and perhaps even open up more if they were speaking their native tongue. Either way, it would be good practice for us to prepare for different scenarios, no matter how unlikely we may think they are, since we would be acting on assumptions. Moreover, we realize that there are some employees who would be extremely difficult for us to help in this sort of program. Some employees may live in areas with poor transit options, which would make public transit unreasonably slow. We realize that the scope of our project is indeed limited; one TDM program cannot be a silver bullet. TMA doesn’t have the power to make decisions that would change the coverage of transit in underserved communities; this is more fit for a long term project that is outside the focus of this project. Moreover, we realize that given how new the PATMA is, we need to be clear and intentional with how we introduce the TMA to employers and employees. We can’t assume interviewees know what a transportation management association is.
For the majority of this week, we will be conducting interviews in downtown Palo Alto. It may be challenging to implement the first few surveys, but those surveys will provide the most valuable information about the quality of our surveys and what areas need to be reworded. Moreover, we will need to adapt our strategies for approaching employees with the survey, contacting the managers/owners, and how to implement the survey comfortably.
As for the Redwood City project, we will continue to work on data analysis. It is challenging to manage both projects at once. Since the trajectory of the Palo Alto project is clearer, we naturally gravitate towards working on that project, so we will need to address this imbalance in attention.