Project Everest

Experiment Results

[EXPERIMENT RESULTS]: Hidden Hunger Malawi - Photo Recognisability: Utility Testing - July 2019

Experiment Post Reference: https://projecteverest.crowdicity.com/post/1351350

 

Lean Phase: Utility Test

 

Purpose: To obtain feedback on the recognisability and clarity of food photographs from locals.

 

Assumption: Having clear and concise photographs to accompany different fruit and vegetable recommendations will make the food survey, and app in general, easier to understand.

 

Time Box: July Weeks 2-3

 

Success Metric: Percentage of positive reactions to accompanying photographs.

 

CRITERIA

Green Light: Assumption is confirmed. Photographs can be used in application to help in identifying which foods are what.

 

Success point: 60% of the customers surveyed had a positive reaction to the accompanying pictures.

 

Orange Light: Between 20% and 60% of customers had a positive reaction. Consider how the photos are being taken and what changes can be made to make them more clear and concise.

 

Red Light: The experiment results contradict the assumption that photographs will help with relaying information in the food survey.

 

Failure Point: 20% or less of customers had a positive reaction.

 

Results

Green light has been achieved. 

 

Data Collection and Representation

 

Photographs of food were taken across several marketplaces (Blantyre Market, Limbe Market and Shoprite). A total of 142 different photographs of food were tested for their recognisability across a sample size of 7 people. The question posed to the participants was “Do you recognise this food?”, with the requirement of a simple Yes or No response. This response was then recorded and placed in the following excel spreadsheet: (https://docs.google.com/spreadsheets/d/1Aw35Taa2djASevsaUV8Pg47Fo8XFVy-aQMeE_JdmzOA/edit#gid=0).

 

Success

The utility test was performed with a total of 143 photographs being showed to 7 people. Out of those 143, 90% of photographs received a positive response with people being able to clearly recognise and often identify each food item.

 

Photographs with a negative response were as follows;

Fried fish in flour

The photograph was not recognised.

Avocado (1st photo)

100% did not recognise as the colouring of avocado was abnormal.

Macadamia Nuts

The framing of the photo created the illusion that the nuts were smaller than they actually are.

Gooseberries

The perspective of camera and placement of the subject made them look like tomatoes.

Grapefruit

Only 1 out of 6 recognised it as something they had seen before, but 100% could not identify it as grapefruit.

Fried Chicken

Less than 60% of people surveyed did not recognise the photograph.

Cashew Nuts

50% recognised, 50% did not.

 

Moving forward

Several photographs need to be retaken to replace the images that were not recognisable. In addition, photographs need to be taken of foods that were not able to be photographed due to seasonal or geographical availability. Another recommendation moving forward is to facilitate a controlled environment with which to set up a photo shoot if the budget allows, again in order to maximise clarity and recognisability of each food item. Photos should be sent to WSU for review.

 

edited on 25th July 2019, 11:07 by Olly Cirocco

Grace Blackford 4 months ago

Status label added: Experiment Results

Reply 0