Assignment 1 Graphs

You might also like

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 2

Comparative Analysis of Methods Used to Obtain

Blood Oxygen Saturation Levels in Infants


The intention of this test is to compare the results given by two different methods of obtaining
an infant’s Blood Oxygen Saturation levels. Each of the methods – C0-Oximetry measurement
and Pulse measurements – were applied at the same time to each of the 60 infants used as
subjects. Each test measures the amount of oxygen present in the infant’s blood as a percentage
of the blood.

Each sets of data give similar results. Firstly, the average result of the CO method is 78.5%, which
is considerably close to the average of the pulse method, which is 75%. It is important to note that
as each are method is a way of obtaining the BOS, it is not possible to determine which is more
‘accurate’, as no true value for the infants BOS levels are known. Therefore, the intention
becomes discovering which method is more reliable. Both have similar ranges, with the pulse
methods beign slightly but not significantly smaller. Complete summaries of each sets of data can
be seen below:

CO Summary:
Min. 1st Qu. Median Mean 3rd Qu. Max.
22.20 71.00 78.50 75.64 84.60 93.50

table 1: Summary of CO test results

Pulse Summary:
Min. 1st Qu. Median Mean 3rd Qu. Max.
24.00 68.00 75.00 72.18 80.00 94.00

table 2: Summary of Pulse test results

Boxplot Comparing CO and Pulse Results

Figure 1 – A Histogram of CO Results, with bar 30


Figure 2 – A Histogram of Pulse results with bar 30

Figure 3 – A Boxplot comparing both methods, CO and Pulse


Figure 1 and Figure 2 both have similar shapes. Their peaks occur at the right side of the graph
and appear to be smooth declines from those points. Figure 1 does state that a higher frequency
of values occurs to the right of the peak while Figure indicates this to the left of the peak. Both
figures have similar spreads, however Figure 1 does evidently indicate the possibility of outliers at
the lower end. While this does occur in Figure 2, the potential outliers are more spread. Figure 1
has all data grouped together while Figure 2’s data is group more sporadically.

Figure 3 is the more interesting graph. The quadrants of CO line up fairly closely to the quadrants
of pulse. However, most notably, the boxplot of pulse indicates far more outliers. While CO does
have 2, the outliers defined by pulse appear more frequently and are grouped together.

From this data I would suggest that both methods of obtaining and infant’s blood oxygen levels
are fairly consistent. Majority of the data for each method, as shown by Figures 1, 2, 3, line up
very closely to each other. The inconsistencies occur when identifying outliers. It is suggested
that utilising the pulse method would be more likely to obtain outliers to the data than using the
CO method. The rest of the data, while different between the two methods, does not
significantly differ between the two methods, therefore I would conclude that both methods are
fairly consistent in obtaining Blood Oxygen Saturation Levels in infants.

You might also like