Deck 23: Laboratory Statistics

Question 1

Which statistical method should be used to answer the following question?
"The published mean for albumin for young adult males is 4.7 mg/dL.A laboratory tested 25 young adult males and found a mean albumin of 4.6 mg/dL ±0.15 mg/dL.Is the mean for the laboratory different from the published mean at the 5% probability level?"

A)t-test
B)unpaired t-test
C)paired t-test
D)linear regression

Answer 1

t-test

Question 2

The F-test is considered a measure of ______ between two sets of data, and the t-test is considered a measure of ______ between two sets of data.______ is/are considered as a parametric comparison.

A)accuracy, precision, the F-test
B)precision, accuracy, both tests
C)accuracy, precision, both tests
D)precision, accuracy, the t-test

Answer 2

precision, accuracy, both tests

Question 3

Suppose a comparison of variation is required between analytical methods, one of which is reported in mg/dL and the other reported in mmol/L.Which of the following measures of variation would be most appropriate?

A)variance
B)standard deviation
C)correlation coefficient
D)coefficient of variation

Answer 3

coefficient of variation

Question 4

Suppose a laboratory is interested in changing from a manual assay for a given analyte to an automated assay for the analyte.Data collected by each of the two methods are unrelated, and the number of observations differs for each dataset.Which statistical method should be selected to determine if a significant difference exists between the means of the two methods?

A)t-test
B)unpaired t-test
C)paired t-test
D)standard error of the mean

Question 5

Considering normally distributed data, the ±1s would contain ______ percent of the data, the ±2s would contain _____ percent of the data, and the ±3s would contain ______ percent of the data.

A)68.26, 95.44, 99.72
B)47.72, 49.86, 68.26
C)13.59, 68.26, 95.44
D)95.44, 99.72, 99.98

Question 6

Degrees of freedom may be defined as:

A)the freedom allowed when selecting data from a dataset.
B)the number of independent observations in a dataset.
C)N - 1.
D)all of the above

Question 7

The two more useful measures of variation in the distribution of data are the variance and standard deviation.Which of the following best explains why standard deviation is the more commonly reported measure of variation?

A)Standard deviation is easier to calculate than variance.
B)Standard deviation provides information on how close together or far apart values are in the sample of data, whereas variance does not provide this information.
C)The units for standard deviation are the same as sample data, whereas the units for variance differ from that of the sample data.
D)Variance is primarily based on the highest and lowest value of the sample data, whereas standard deviation is based on all sample data.

Question 8

Following comparison of a new method for measuring calcium with a reference method, the equation of the line to describe the relationship between two methods was determined to be y = 1.044x - 0.022, with r = 0.9982.Given this information, a calcium level of 5 mmol/L determined by the reference method is equal to what calcium level by the new method?

A)4.8 mmol/L
B)4.9 mmol/L
C)5.0 mmol/L
D)5.2 mmol/L

Question 9

In the evaluation of how well data fit a normal distribution, ______ serves as a measure of how steep or flat the distribution is when compared to a true normal distribution.______ serves as a measure of the asymmetry of the data distribution.For data that reasonably follow a Gaussian distribution, skewness and kurtosis coefficients should be between ___ ___.

A)skewness, kurtosis, 0 and 1
B)skewness, kurtosis, -1 and 1
C)kurtosis, skewness, 0 and 1
D)kurtosis, skewness, -1 and 1

Question 10

If data are non-Gaussian distributed, steps are typically taken to transform data by taking the logarithm of the data to produce a Gaussian distribution.Why are laboratories interested in data existing in a Gaussian distribution?

A)NCCLS standards require that only Gaussian-distributed data be utilized in laboratory statistics and quality control.
B)Transformation to a Gaussian distribution will allow the utilization of applications useful to the clinical laboratory.
C)Transformation to a Gaussian distribution will correct data points causing the initial data to not be Gaussian distributed.
D)All of the above

Question 11

______ describes the reproducibility of a method, and ______ describes the ability of the method to obtain the true or correct result.The ______ the standard deviation, the better the precision of the method.(answer format: blank 1, blank 2, blank 3)

A)Accuracy, precision, smaller
B)Precision, accuracy, smaller
C)Accuracy, precision, larger
D)Precision, accuracy, larger

Question 12

Suppose there is a set of normally distributed data from which a normal reference interval for health is to be determined.If the reference interval is set as 0 ± 2s , what percentage of normal healthy individuals will fall outside the normal reference interval for health?

A)0 (none)
B)0.28%
C)2.14%
D)4.56%

Question 13

The difference between parametric and nonparametric statistics is based on:

A)the assumption of a normal (Gaussian) distribution of the data.
B)the method used to consider outliers.
C)the assumption that only systematic error is present.
D)the equivalency of the mean, median, and mode.

Question 14

Which statistical method should be selected in the following situation?
"Suppose a laboratory is interested in determining the prognostic ability of selected laboratory tests to predict patient outcomes following severe head trauma, based on a comparison of survivors versus non-survivors at predetermined times following the head trauma."

A)t-test
B)unpaired t-test
C)paired t-test
D)chi-square analysis

Question 15

Which of the following correctly describes the difference between a correlation coefficient (r) of +1 and -1?

A)the slope of the regression line
B)the degree of correlation between the variables
C)the direct/inverse relationship between the variables
D)both a and c

Question 16

Suppose a large laboratory is interested in a comparison of the means for glucose run on two different analyzers in the laboratory.To conduct this comparison, the same specimen is analyzed on each analyzer, and data are collected.Which statistical method should be selected to perform this comparison?

A)t-test
B)unpaired t-test
C)paired t-test
D)F-test

Question 17

Which of the following measures of central tendency identifies the true central tendency for all types of distributions?

A)mean
B)median
C)mode
D)frequency