Example: Patient Glucose Levels (mg/dL): __ y Bt x tetx, + Patient 1: 90 oe * Patient 2: 110 n i=1 * Patient 3:95 90+110+95+120+100 * Patient 4: 120 5 * Patient 5: 100 ® =103 Interpretation: The mean glucose level of these patients is 103 mg/dL.Median * Data in the middle after data has been ranked from least to greatest, divides the distribution of data points into upper and lower halves, or the 50* percentile, where 50% of values are greater and 50% are lower. * If the number of observations is odd. eo... -Median is the middlemost observation in an array * If the number of observations is even. -Median is the average of the two middlemost observations in an array.Weight Weight 69 56 a 7 58 58 Since the number of observations is 84 a odd (ie., 11), the median is the 7 a middlemost value (i.e., 6) arranged in an array — 79 oe o 7 z ai Interpretation: 50% of students have t weights < 69 kilos, while the other 81 Ls 50% have weights > 69 kilos. 68 84 Arrange the observations in an array (i.e, lowest to highest)69 +69 = Median 7 Arrange the observations in an Weight 6 Weight 9 56 a [57 58 38 84 6 57 56 79 79 9 81 83 83 81 34 array (.e., lowest to highest) Since the number of observations is even (i.e., 10), the median is the average of the 5 and 6 values arranged in an array Interpretation: 50% of students have weights > 69 kilos, while the other 50% have weights < 69 kilos.Mode * The most frequently occurring or most prevalent value in the data set/the point at which the greatest number of values occur. -Can be absent or can be multiple (i.e., multi-modal) * Computation is generally not needed.Example: Number of Patients Diagnosed with Diabetes in Different Age Groups: * Age Group 20-30: 5 patients ax — vod * Age Group 31-40: 10 patients Cini wo dal * Age Group 41-50: 10 patients * Age Group 51-60: 7 patients * Age Group 61-70: 8 patients Identifying the Mode: In this case, the age groups 31-40 and 41-50 both have 10 patients, which is the highest frequency. 2 dal 414 (9) VF 51? A) 24 dioMeasures of Dispersion/Spread * Gives an indication of the tendency of observations (or values) to clump together * Computed for quantitative variables only * Measures of Dispersion - Range - Standard deviation - Coefficient of variation (CV) - VarianceRange * The range is the difference between the largest and smallest value in a set of observations (or values) * Can be expressed as a single number, but it is preferred to present the range as a pair of numbers (i.e., minimum, maximum) * Does not tell anything about observations between the two extreme observationsExample: Patient Glucose Levels (mg/dL} Patient 1:90 Patient 2: 110 Patient 3: 95 Patient 4: 120 Patient 5: 100 Range= Highest value- Lowest value Range = 120-90 Range = 30 mg/dL Interpretatior + The glucose level of the patients range from 90 mg/dL to 120 mg/dl (preferred) + The range of the patient's glucose levels is 30 mg/dLVariance + Measures of dispersion that relates the scatter of observations (or values) to the mean - Values lie close to the mean - Less dispersion - Values are scattered about the mean - More dispersion(@i-%)] (xi - 2) ad 9.4 124 134 23d 144 84 14 124 104 254 92.14 158.74 179.56, 184.94 213.14 70.54 254 153.74 108.14 Zhe -X)? n-1 1,166.20 10-1 1,166.20 9 129.58 kilos?Standard Deviation * Obtained by taking the square root of the variance * The unit of the standard deviation is the same as that of the original set of observations. * Standard deviation index - Most appropriate test when comparing laboratory’s monthly mean to its peer group to determine if bias is present#[ x | * [@-d[ai-m s 1 | 69 [706 | -16 | 256 2[ 4 | 706 | -96 | 9216 3 [58 [706 | -126 | 158.76 4 | | 706 | 13.4 | 179.56 5 | 87 | 706 | -136 | 184.96 6 | 56 | 706 | 14.6 | 213.16 7 [79 [706 | 84 | 7056 8 | 6 | 706 | -16 | 256 9 | 83 [706 | 12.4 | 153.76 v0 [ 81 | 706 | 104 | 108.16 $= 11.38 kilosExample: Assume that the laboratory mean is 9, the consensus group mean is 8, and the consensus group standard deviation is 2. SDI = (9 - 8)/2 = 0.5 Interpretation: The positive value of SDI indicates that the laboratory mean is greater than the consensus group mean. The value 0.5 indicates that the means are 0.5 standard deviations apart.Coefficient of Variation * Expresses the standard deviation as a percentage of the mean * Useful when comparing the dispersion: - Between two sets of data with different units - Between two sets of data with the same units but significantly different means * Enables the medical technologist to compare standard deviations for various units and concentrations expressed as percentages. __ Standard deviation cv= Mean x 100%9 58 84 57 56 79 6 83 ai Weight fin kos) Height (centimeters) — 2 = Gi-#) GiB HF Gt) Gt-#— Vorigne = FX 100 706-16 256157 1637 4409 706 96 9216 181 1637 16129 11.38 kg 706 126 15876 183. 1637 11649 Yeeine = rpap pq * 100 = 16.12% 706 134 17956 176 1637 151.29 : 706 136 10496 162 1637 289 706 "146 21316 175.1637 12769 CVggigne = £100 70.6 84 70.56 153 163.7 114.49 — 3 706 16 256 163 1637-07049 706 124 15376 172 1637 836889 cy,,,,,, = 10.08em 70.6 10.4 108.16 17S_ 163.7 11.3 127.469 weit“ T63.70 cm i= 12988 si= 10187 so 1138 s= 1008 latempretation: Since the coefficient of variation of weight is greater than the coefficient of variation of height, weight is more heterogeneous (or more dispersed) than heightMeasure of shape * Refers to how the distribution of data is characterized, particularly in terms of its symmetry and peakedness. Common measures of shape include: 1.Skewness: This indicates the degree of asymmetry in a distribution. 1. Positive skew: Tail on the right side is longer or fatter, indicating that most values are concentrated on the left. 2. Negative skew: Tail on the left side is longer or fatter, indicating that most values are concentrated on the right. 2.Kurtosis: This measures the “tailedness" of the distribution, indicating how much of the data is in the tails versus the center. 1. High kurtosis: Indicates more data in the tails (leptokurtic), suggesting more outliers. 2. Low kurtosis: Indicates less data in the tails (platykurtic), suggesting fewer outliers.Skewness Kurtosis Positive www.statisticalaid.comGaussian Distribution * Spread of data in which the elements are distributed symmetrically around the mean with most values to close to the center. * Confidence Interval- interval that is computed to include a parameter such as the mean with stateg probability that the true value falls into the interval. - 68.3% is between +/- 1SD - 95.8% is between +/- 2SD -99.7% is between +/- 3SDInferential Statistics * Used to draw conclusions . . eee (interference) regarding SD and ie =-je22 means of two sets of data or more. : . see e + ° ee ceInferential Statistics * T-Test Used to determine if there is a significant difference between two sets of data or sample means or means of two methods. Can either be dependent or independent * F-Test Used to compare the magnitude of the new method’s imprecision to the magnitude of the reference method’s imprecision. * Analysis of Vai@ance (ANOVA) Allows for the examination of possible differences between three or more independent groups.Inferential Statistics * Correlation Aterm that refers to the relationship or correlation of two quantitative variables. * Linear Regression Used to evaluate the performance of a new method in comparison to an established (reference) method