Mean Spree Headband Temperature: 102.9 degrees Celsius
Mean Gold Standard Temperature: 97.6 degrees Celsius
Mean Spree Headband Heart Rate: 92.82 beats per minute
Mean Gold Standard Heart Rate: 92.52 beats per minute
The standard deviation is shown through the error bars on the graphs.
Figure 1: Graph of the heartbeat results obtained from the gold standard and the spree. The mean heart rate from the gold standard was observed to be 92.52 beats per minute and the mean heart rate from the spree headband was observed to be 92.82 beats per minute.
Figure 2: Graph of the temperature results obtained from the gold standard and the spree. The mean temperature from the gold standard was observed to be 97.6 degrees Celsius and the mean temperature from the spree headband was observed to be 102.9 degrees Celsius.
Inferential Statistics
T-tests were used rather than ANOVA tests, because in each group only two sets of data were being compared.
Figure 3: T-test results for heart rate results from the spree headband and the gold standard.
Figure 4: T-test results for temperature results form the spree headband and the gold standard. Small values for the p-value mean that the correlation between the variables under investigation rarely occurs.The T-test for temperature shows a extremely small p-value also implying that reliable and valid results for temperature cannot be obtained from the spree headband.
Analysis
By comparing the values of heart rate between the gold standard and the spree headband we find that there is not a statistical correlation between the two. The p-value found in the T-test showed that there might be no relationship between the two measurements but we cannot be sure since the value is less than 0.1. The gold standard is assumed to be as close to the true value as possible. The spree headband produces a measurement that is somewhat close to the true value but not acceptable enough.
Comparing the values of heart rate between the gold standard and the spree headband for measuring temperature we found a correlation between the two. The p-value was much lower than the accepted p-value that states that there is a correlation between the two measurements. The p-value obtained from the T-test between the spree and the gold standard was 4.06*10^-45 which practically insures that the spree headband will not get the same results as the gold standard.The low p-value obtained also implies that the correlation between the spree and the gold standard rarely occurs mainly due to random sampling. This then makes it difficult for the same results to be obtained in the event of repeating the experiment.
Summary/Discussion
The variance between the Spree headband's measurements of body temperature and heart rate compared to the gold standard is too high to be acceptable. Though the values from the spree headband were highly reliable (similar measurements obtained), there are not valid as there are far off from the acceptable values obtained from the gold standard. Our group thus came to the conclusion that modifications need to be made to the Spree headband in order for it to measure body temperature correctly.
Problem 1: The headband needs to be on really tight to ensure it does not fall off the user, which causes discomfort
Problem 2: The headband is difficult to connect to the phone app
Problem 3: The band needs to be more appealing to wear
Problem 4: The app does not have a usable temperature gauge because it does not measure in degrees of Fahrenheit or Celsius
Problem 5: The heart rate is inaccurate.
Solution 1: We can create a more suitable design for the Spree band with memory foam inside the headband so that the headband can take the shape of the person's head and remember the shape of the person's head.
Solution 2: We can redesign the wireless connectors in the Spree headband so that they will generate a stronger wireless connection with the phone application. Also, we could have the spree band bluetooth names custom so it is easier for the user to identify which band to connect to.
Solution 3: We can redesign the Spree headband with different materials such as the airy materials used to make running shoes. We want the headband to be light so that the person wearing it will be comfortable wearing the headband often while working out.
Solution 4: A major design flaw inherent in the spree headband was its inability to take accurate readings of internal body temperature, mainly due to the fact that it takes temperature readings from the skin, which does not always reflect internal temperature. A modification that may be able to remedy this issue would be to make an internal device, possibly one that could fit in the mouth much in the way a retainer does. This would allow for much more accurate temperature readings and allow temperature to display as an exact number rather than a temperature range.
Solution 5: Another flaw in the headband was the inaccuracies in measuring heart rate. This could be due to its position on the body, as a headband, the spree headband can be subjected to jostling and shifting, which could throw off readings. One potential fix for this problem would be to do away with the headband design entirely and follow the example of previous devices which strapped to the chest with a heart rate monitor positioned directly over the heart itself.
LAB 3B WRITE-UP
Target Population and Need
Target Population: Athletes
Need: Our device, Gyro Guard, can detect concussions early so that athletes can treat the symptoms of a severe concussion before the symptoms become too fatal. Millions of athletes suffer from concussions every year without being diagnosed of a concussion.