Two scattter diagrams, each scatter diagram showing three points with confidence intervals for each point.

In the first scatter diagram the horizontal axis is labeled "Study number" and labeled 1, 2, and 3. The vertical axis is labeled "Difference in proportion healed" and numbered -1.5, -1, -0.5, 0, 0.5, 1, and 1.5. The first point is at study number = 1, difference in proportion healed = 0.15. The confidence interval line goes vertically from -0.1 to 0.4. The second point is at study number = 2, difference in proportion healed = 0.3. The confidence interval line goes vertically from 0.15 to 0.45. The third point is at study number = 3, difference in proportion healed = 0.2. The confidence interval line goes vertically from -0.1 to 0.5. The areas above difference = 1 and below difference = -1 are shaded and labeled "Impossible values".

In the second scatter diagram the horizontal axis is labeled "Study number" and labeled 1, 2, and 3. The vertical axis is labeled "Number needed to treat" and numbered -20, -10, 0, 10, and 20. The first point is at study number = 1, Number needed to treat = 8. The confidence interval line goes vertically from 3 to end in an upward pointing arrow above 20. Another line extends vertically downwards from Number needed to treat = -17 to a downward pointing arrow below -20. The second point is at study number = 2, Number needed to treat = 4. The confidence interval line goes vertically from 2 to 12 The third point is at study number = 3, Number needed to treat = 5. The confidence interval line goes vertically from 3 to end in an upward pointing arrow above 20. Another line extends vertically downwards from Number needed to treat = -9 to a downward pointing arrow below -20. The area between Number needed to treat = -1 and Number needed to treat = 1 is shaded and labeled "Impossible values".

Back to Clinical Biostatistics: Analyses for qualitative data.

This page maintained by Martin Bland.

Last updated: 9 August, 2006.