# Limits Of Agreement Spss

“A Bland-Altman analysis was used to assess the degree of agreement between the two methods, to compare the new technique with the established one.” Note that the 95% limit values for differences between the measured values are about plus or minus 75, and the graph shows some differences outside or near border lines, i.e. differences in the vicinity of ±75 (the actual limit values are -75.4 and 79.6). Bland and Altman (1986) argue that such important differences are clinically important (a decision that was not based on a p value) and therefore conclude that the two devices are not sufficiently compliant to be used interchangeably. It is important to note that this lack of agreement is not evident in the identity conspiracy. The following steps are used to calculate the values required for this graph. Bland, J.M., Altman, D.G. (1986). Statistical methods to assess the agreement between two methods of clinical measurement. Lancet, 327 (8476), 307-310. While the SPSS does not have specially designed equipment to create Bland-Altman diagrams, they can be produced in SPSS using the graphic editor. If the measurements are recorded in variables A and B, the difference between A and B can be calculated and stored as a new variable (e.g.B. DIFF) in the Transform>Compute with destination variable and “A-B” (without quotation marks) as a numerical expression. Similarly, the average value of the two measures (z.B.

MMEAN) can be calculated in the Transform>Compute with the target variable of MSAN and ” (A-B)/2″ as a numerical expression. To print statistics that can be described on LA DIFF, as well as a test to determine if the DIFF has an average value of 0, run the T-type test method (Analysi>Compare Means->One-Sample T Test) with diff in the “Test Variable (s)” field. Enter a “0” in the “Test Values” field. The output of the one-sample T test includes the average and standard deviation of the DIFF, as well as the typical error of the average value, confidence intervals for the average value (default: 95%) and the level of meaning for the test that the average value of the DIFF is equal to 0. . · Bland-Altman Diagram: A dispersal diagram with variable averages, represented on the horizontal axis, and the differences represented on the vertical axis that indicates the amount of discrepancies between the two ratios (on differences) and allows you to see how this disagreement relates to the magnitude of the measurements.

This entry was posted in Geen categorie. Bookmark the permalink.