Project Everest

Adopted Experiment

[Jan 19] Data Communication Incentives (Channels) - FarmEd Timor-Leste

by
Marina E.M. Smith
Marina E.M. Smith | 9 months ago | in FarmEd - Timor-Leste

Does Providing Incentive for Communication of Data Increase the Success Rate for Sensor Data Reporting?

 

Introduction:

An earlier case study, which was run from July through to December 2018, attempted to test several aspects of sensor usage; the ability of farmers to utilise the sensor correctly, read the results, record the results, and finally, communicate these results to the FarmEd team through the communication platform of Facebook.

The results of this study (see attached PDF) confirmed that farmers are fully capable of, and keen to utilise the sensor correctly, read the results, and record the results, however, none of the six participating farmers achieved success in communicating the data. When members of the December FarmEd team visited participants, it was confirmed that the measurements had been taken, and the data recorded by hand into notebooks, despite the lack of facebook communication.

This experiment aims to verify if a lack of incentive was responsible for this outcome. It is possible that as no clear reason for communicating the data was provided to farmers, the investment of time and internet towards this task was not viewed as a worthwhile expenditure.

The importance of incentive will be assessed through comparison of this experiments’ results with those of the earlier sensor case study. As no advice was offered as an incentive for reporting results in this earlier experiment, it’s success rates can be used as a control against those of this ‘incentivised’ experiment.

The incentive provided for this experiment is the provision of personalised farming advice, designed on the basis of the communicated sensor data at the successful completion of the experiment. It will be emphasised throughout interactions with the farmers that the personalised agricultural advice will only be provided at the conclusion of the experiment, to individuals who have successfully completed their required tasks.

This advice will be provided to farmers digitally, however, the February FarmEd team will visit each farmer to confirm that the advice has been received and understood.

 

Lean Phase:

Channels.

 

Assumption:

Farmers require the incentive of clear personal gain to justify expenditure of time to consistently communicate sensor data to the FarmEd team/Project Everest.

 

Time Box:

Four weeks.

 

Success Metric:

Green Light- If the success point is reached, it has confirmed that farmers place value in the concept of receiving individualised advice based on the reported sensor data. This indicates farmers are likely to engage with this aspect of the FarmEd application. As such, a positive outcome to this experiment would confirm that farmers will enter sensor data into the app.

Success point- 70% of farmers successfully report the data, where ‘success’ is defined as communicating the sensor data in full for at least three of the four weeks. Of the 30% who fail to successfully report their data, the majority can reasonably be assumed to have failed due to factors unrelated to incentive.

Orange Light- Consider the manner in which incentive has been presented. Consider whether the communication channels, and the variables entailed in these are too technologically complex, or could be presented in a different manner which minimises the time/effort investment by the farmer in communicating data. Also consider such factors as whether it is an accurate reflection of the service we intend to provide, whether an example of advice, eg an advice report from another farm support the concept of the value, or even a list of what exactly advice will be provided on, clarify the purpose for sensor data reporting?

Failure Point - less than or equal to 30% of farmers successfully report the data. This indicates that there is insufficient incentive for farmers to invest time in communicating sensor data.

Red Light- Consider whether the advice we are providing is an adequate reflection of the individualised service provided by the app. If it is, then does this indicate a lack of value perceived by farmers in the individualised service? Or are there alternate explanations for this outcome?

 

Experiment build:

Method

  1. Make a template for facebook data reporting, as well as a facebook message which can be sent out as a reminder for farmers each week.

  2. A spreadsheet will be made where the weekly data reports (pH and moisture with five measurements of each, per farm area, per week) can be recorded. An accompanying document will also be made to record information about each participating farmer, as well as any questions, feedback, difficulties encountered or otherwise potentially relevant information.   

  3. Farmers will be recruited to participate through the channel of community meetings. In these meetings, we will present the sensors for sale as part of the presentation, and include information on the experiment itself. Any farmer who wishes to purchase a sensor for US$6 will be informed of the data reporting process in individual conversation, in an attempt to encourage their participation. The goal is to recruit ten farmers.

  4. Once a farmer has agreed to participate, an individual farm visit will be arranged, wherein team members will visit the farm, teach the farmer how to operate the sensor, record the data, and send the data through facebook.

  5. A facebook message reminder will be sent to all participants every Monday.

  6. When farmers send their data, it will be recorded, and the success/failure table which is associated with each participant, located in the google doc, will be filled out.

  7. FarmEd team will analyse the data obtained, and utilising the ag guide, will construct basic agricultural advice for each farmer.

  8. After four weeks of data collection, farmers will be contacted with a brief message congratulating them on their completion of the trial, and provide them with their personalised advice digitally. Farmers who successfully completed the trial will receive more in-depth advice. They will also be informed of an impending visit from the February FarmEd team, who will confirm that the advice has been received, as well as collecting farmer feedback.

 

Resources Required

10 sensors 

Experiment general information google document

Experiment results recording spreadsheet

Access to agricultural information to provide accurate advice

 

Potential Problems/Blocks and Solutions

Block: Explaining the purpose, and importance of communicating data to participating farmers

Solution: Perhaps provide written materials as well as verbal explanations, to aid in understanding but also for later reference

 

Block: Maintaining the record of data and other information.

Solution: The tasks required for this will need to be detailed in the January 2019 handover document, to ensure the February 2019 team are clear on the requirements

 

Block: Revisiting participants after the experiment to obtain feedback.

Solution: The location of participating farmers will be marked clearly on the FarmEd engagement map, and where possible, photos of the entrance to each property will also be provided, where possible, for future ease of location.


February Follow-up Requirements

  • Provide farmers with advice on completion of the experiment
  • Complete the survey/questionnaire for participating farmers
  • Publish the results of the experiment on crowdicity
  • Analyse the outcomes of the experiment, and determine how to proceed

 

Notes:

  • Aim to find ten farmers who wish to actively participate
  • Aim to recruit farmers from 3 different regions
edited on 17th January 2019, 02:01 by Marina E.M. Smith

Harry Telford 9 months ago

Status label added: Experiment adopted

Reply 0

Julia Marks 9 months ago

Hi Marina.

I love the idea of continuing the sensor case study as it will provide us with great information about how the farmers will interact with the application.

I'm just curious as to what you are using as the 'incentive' in this experiment to entice farmers to send data. Is it that you will provide them with agricultural advice? If this is the case, then that is very similar to the initial agreement made in July and what was carried out in December.

Have you considered any other incentives?

Reply 0

Marina E.M. Smith 9 months ago

Hi Julia!

Thank you for your comment! Yes the incentive for this experiment is the provision of personalised agricultural advice in response to each farmers' successful provision of the sensor data to the FarmEd team via the Facebook page.

From what we understood of the case study from July-December, there was no such incentive provided to participants, to encourage consistent data communication. However, if we are incorrect about this, we would very much appreciate it if you could point us in the direction of some more information. :)

Reply 0