The T-Test for temperature shows that the data is not significant because the p-value is greater than 0.5
Heart Rate P-Value:0.5
The T-Test for heart rate shows that the data is not significant because the p-value is greater than 0.5.
Summary/Discussion
The start up company, Suntronic has asked us to test one of the latest heart rate and temperature monitors for smart phones by Spree Sports. To test the validity of the Spree Sports head band and app, temperature and heart rate readings from the device were compared to temperature and heart rate readings from oral thermometers and pulse oximeters. The readings given by the oral thermometers and the pulse oximeters are considered the gold standard.
Upon comparing the results from the Spree Sports sensor to the thermometers and pulse oximeters, several design flaws were noted.
The initial flaw noticed in the design was the bluetooth connection. The Bluetooth system of spree doesn’t work very well, so there was difficulty syncing the Spree sensor to the electronic applications on the smart phones. If the bluetooth connection cannot be improved, it is recommended that each Spree sensor come with a code that can be manually put into the application on the phone.
For the high price of the Spree headband, several factors are quite unsatisfactory. The app, for instance, has very few features. The readings for the temperature in the Spree app is based on a very ambiguous scale that doesn’t have any concrete numeric values. Instead, it gives temperature readings on a scale system from levels 1 to 4. The intervals for measuring temperature are not established and therefore an accurate reading of temperature cannot be taken from the application. It is advised that the readings detected give actual temperatures for more precise readings. Also, the Spree app is only available on iOS operating systems which puts the company at a huge disadvantage by discounting a large amount of the potential target market. This can easily be fixed by the company by making the app available to android users.
In terms of design, the Spree headband contains critical flaws. The placement of the Spree sensor only detects the external temperature, decreasing the accuracy of the temperatures given. The core temperature being maintained during exercise alters the skin temperature that is detected (on a limiting scale) by the Spree sensor. Because an internal temperature would be difficult to measure, the easiest fix for this feature is to move the location of the Spree sensor closer to the core of the body. Instead of the head band, it is recommended that the sensor be placed in a waist band or a band that can be secured around the abdomen. Also, the sensor does not fit snuggly into the headband, so it falls out very easily. An easy fix to this flaw would be to create a more secure compartment in the band for the sensor.
From the values obtained from the inferential statistics, in this case the T-test, for heart rate calculations, the spree and the gold standard groups were significant (percent error=1.6886E-5). This happened because the reading that the spree application gave for the heart rate values was fairly accurate, providing three significant figures. This fact also contributed in a Pearson correlation value closer to 1 (Pearson's R=0.984), as the measurements from the spree application and the gold standard were very similar between each other.
However, the values obtained from the T-test for the temperature calculations were not significant (percent error=7.6187E-15). The Pearson correlation was far from 1 (Pearson's R=-0.02886) because the temperature measurements given by the spree were different from the gold standard temperatures. This indicates that the Spree sensor and app have critical flaws that need to be fixed in terms of temperature readings.
BME 100 Fall 2014
