BME100 s2017:Group9 W8AM L3
BME 100 Spring 2017 | Home People Lab Write-Up 1 | Lab Write-Up 2 | Lab Write-Up 3 Lab Write-Up 4 | Lab Write-Up 5 | Lab Write-Up 6 Course Logistics For Instructors Photos Wiki Editing Help | ||||||
OUR TEAMLAB 3 WRITE-UPDescriptive Stats and Graph
Standard Deviation: 21.66 Spree Average Heart Rate: 99.146 Standard Deviation: 24.33 Correlation Coefficient: 0.78 Gold Standard Average Temp: 96.65 Standard Deviation: 1.919 Spree Average Temp: 95.53 Standard Deviation: 0.869 Correlation Coefficient: 0.193
Inferential StatsTemperature Spree Headband Average temperature 95.53 Spree Headband Standard Deviation temperature 0.869034 Gold Standard Standard Deviation temperature 1.919 Gold standard average temperature 96.647 Paired t test P value <0.0001 Significantly different (P < 0.05)? yes two-tailed P value t=10.3 df=323
How big is the difference? -1.116 Mean of differences 1.952 SD of differences 0.1084 SEM of differences -1.33 to -0.903 95% confidence interval 0.2471 R squared (partial eta squared) Number of pairs: 324 Correlation coefficient (r): .193
Gold standard average heart rate 98.76 Spree headband average heart rate 99.146 Gold standard standard deviation heart rate 21.66 Spree headband standard deviation 24.3314 Paired t test 0.6708 P value not significant Significantly different (P < 0.05)? No two-tailed P value t=0.4255 df=299
How big is the difference? 0.3793 Mean of differences 15.44 SD of differences 0.8915 SEM of differences -1.375 to 2.134 95% confidence interval 0.0006052 R squared (partial eta squared) Number of pairs: 300 Correlation coefficient (r): 0.78 Design Flaws and RecommendationsThe Spree headband does not appear to effectively measure body temperature as the paired t-test between the Spree headband and an oral thermometer gave a p-value of <0.0001 This p value is <.05 which is statistically different from the values provided by the gold standard oral thermometer. The R correlation is .1928, which is extremely low, showing that the spree headband is almost completely inaccurate. On the other hand, the Spree headband measures the heart rate fairly accurately compared to the gold standard since the p-value was .6708, which is >.05, meaning it is statistically similar. The R correlation is .78, which is close to 1, which would mean a perfect correction, therefore showing that the device could use improvements, but is fairly accurate. The suggestion for the Spree Headband would be to completely scrap their current temperature since it is does not reflect the true temperature of the body, or to stop advertising that the device can read temperature.
Experimental Design of Own DeviceTo test our design we would start with a large sample size of at minimum 300 women within the age range 18-80. A simple random sample would be utilized to select the subjects. Once a month, for a year, each woman would be tested for breast cancer, or differences in tissue density, with a mammogram as well as our device. Those who are reading the results will not know which device they are reading from to avoid bias. If the mammogram detects any discrepancies, our device should as well, making the mammogram the gold standard. We would compare the accuracy of the mammogram to detect density discrepancies to the ultrasound patch's ability to detect density discrepancies. Our device should match the accuracy of a mammogram or be more accurate. To analyze the data, we would use a paired T-Test because we would be comparing two data entries for each person. |