You are viewing documentation for the old version 2.30 of Analyse-it. If you are using version 3.00 or later we recommend you go to the Bland-Altman difference plot.

Altman-Bland

Altman bland test determines the agreement between two variables.

The requirements of the test are:

  • Two variables measured on a continuous scale.
  • Replicates can be observed for each variable, though the two variables need not have the same number of replicates.
  • Each cases must have the same number of replicates, otherwise the entire case will be listwise deleted.


Arranging the dataset

Data in existing Excel worksheets can be used and should be arranged in the List dataset layout. The dataset must contain at least two continuous scale variables. If replicates are observed then a List dataset with grouped variables layout should be used to arrange the replicates for each variable.

When entering new data we recommend using New Dataset to create a new 2 variables dataset ready for data entry.

Using the test

To start the test:

  1. Excel 2007:
    Select any cell in the range containing the dataset to analyse, then click Agreement on the Analyse-it tab, then click Bland-Altman.  
  2. Excel 97, 2000, 2002 & 2003:
    Select any cell in the range containing the dataset to analyse, then click Analyse on the Analyse-it toolbar, click Agreement then clickBland-Altman.

  3. Click Variable X and Variable Y and select the variables.
  4. Click OK to run the test.

The report shows the number of cases analysed, and, if applicable, how many cases were excluded due to missing values. The number of replicates and repeatability SD/CV (depending on the Plot option, see below) of each variable is shown.

The scatter plot (see below) shows the observations of X plotted against Y. The Use replicates option determines how replicates for each variable, if available, are plotted. 

(click to enlarge)

Understanding and configuring the difference plot

The difference plot shows the difference between the variables against the average of the variables. The scatter of the differences around the zero line should be constant, if the scatter shows a funnel effect with the differences been larger for higher values you should consider using the % difference plot to try achieve a constant scatter.

To change the difference plot:

  1. If the Bland-Altman agreement dialog box is not visible click Edit on the Analyse-it tab/toolbar.
  2. Click Plot and select Difference to plot absolute differences, or Difference as % to plot the differences as a percentage of the average.
  3. Tick Histogram of differences to show a histogram of the distribution of the differences.
  4. Click OK.

The difference plot (see below) is shown beneath the scatter plot.

(click to enlarge)

A histogram of differences (see below), with a normal curve overlay to assist in judging whether the differences are normally distributed, is shown beneath the difference plot.

(click to enlarge)

Determining and plotting bias

Bias is the average difference between the variables and should ideally be zero. Although, correcting for a non-zero bias is a simple matter of subtracting the bias from the variable.

The bias and a confidence interval are shown. A hypothesis test is also shown. The p-value is the probability of rejecting the null hypothesis, that the bias is equal to zero, when it is in fact true. A significant p-value implies that the bias is difference from zero.

The bias and confidence interval can also be overlaid on the difference plot.

To show bias overlay:

  1. If the Bland-Altman agreement dialog box is not visible click Edit on the Analyse-it tab/toolbar.
  2. Click Overlay and select with Biasto show the bias, or Bias + CI to show bias with a confidence interval on the difference plot.
  3. Enter Confidence level to calculate around the bias, as a percentage between 50 and 100, excluding the % sign.
  4. Click OK.

The difference plot (see below) shows the bias and confidence interval.

(click to enlarge)

Determining and plotting Limits of agreement

Limits of agreement measure the agreement between the variables. They represent a range in which a given percentage of the differences lie. Large limits of agreement show poor agreement between the variables. In some cases the poor agreement maybe due to poor repeatability (see below).

The limits of agreement and a confidence interval for each are shown.

The limits of agreement and confidence interval can also be overlaid on the difference plot.

To show the limits of agreement overlay:

  1. If the Bland-Altman agreement dialog box is not visible click Edit on the Analyse-it tab/toolbar.
  2. Click Overlay and select with Bias + Limits of agreementto show limits of agreement, or Bias + Limits of agreement + CI to show limits of agreement with a confidence interval on the difference plot.
  3. Enter Limits of agreement, as a percentage between 50 and 100 excluding the % sign.
  4. If either variable is observed in replicate, click SD differences and select Between mean measurements to calculate the SD from the differences between the mean of the variables, Between single measurements ... TODO
  5. Click OK.

The difference plot (see below) shows bias, limits of agreement, and confidence intervals.

(click to enlarge)

Examining Repeatability

When a variable is observed in replicate we can measure the repeatability. The repeatability as a SD or CV is shown along with the repeatability coefficient which shows the limit which we would expect the differences between two measurements to lie. If the variable shows poor repeatability we would expect the limits of agreement to be poor.

The repeatability can also be shown as a plot.

To show repeatability plots:

  1. If the Bland-Altman agreement dialog box is not visible click Edit on the Analyse-it tab/toolbar.
  2. Tick Repeatability plot.
  3. Click OK.

Repeatability plots (see below) are shown for the variables measured in replicate.

(click to enlarge)

References to further reading

    1. Measurement in Medicine: the Analysis of Method Comparison Studies
      D.G. Altman, J.M. Bland, The Statistician 32 1987; 307-317.
    2. Statistical Methods for Assessing Agreement Between Two Methods of Clinical Measurement
      J.M. Bland, D.G. Altman, The Lancet February 8 1986; 307-310.
    3. Measurement Agreement in Method Comparison Studies
      J.M. Bland, D.G. Altman, Statistical Methods in Medical Research 1998; 8: 135-160.
    4. Comparing Methods of Measurement: Why plotting difference against standard method is misleading
      J.M. Bland, D.G. Altman, The Lancet 1995; 346: 1085-87.
    5. Agreement Between Methods of Measurement with Multiple Observations Per Individual.
      Bland J. Martin, Altman Douglas G, Journal of Biopharmaceutical Statistics; 17 2007: 571-582.

(click to enlarge)