NUSHJC Team 1

Allocate time along with area

• Consider what the participant’s daily schedule is like

• So that you collect all the data you need and not a bunch of biased data.

• In SGC the data was still relatively spread because we were asked to collect data from both AM and PM

 

Make it clear that they should collect data from all parts of the park equally

Consider implementing an incentive scheme to get people to collect data

• Keep in mind that people have difficulty in finding time

• Incentivise well collected data

http://www.cs.iit.edu/~xli/paper/Journal/TPDS2297112.pdf

http://www.joe.org/joe/2014october/iw2.php

 

 

The sensors given to participants need to be good

• A lot of people's time will be used and there may be a big investment in terms of infrastructure and incentives and so this needs to be justified with good data

• Cleaning the data well can be quite expensive

 

Participants need to be very sure on how to use the equipment

• This will prevent incorrect usage of sensors, giving more reliable data

• This is not only for sensors, but also while dealing with the data and software

• The instructors have to emphasize the importance of certain common pitfalls so that participants don't render their data and efforts useless because of a simple mistake 

 

One possible method to ensure that the data collected is over the same time period across the entire park is to organise data collection events. It could be a data collection drive style event where people can volunteer to spend a day at the park collecting data. Apart from this giving us good data that we can use to compare different parts of the park more reliably, we can also help the volunteers with hardware issues they face by stationing an expert at the venue. But of course the event should not be too large scale that it disrupts the park environment and creates a buzz in the park.