It is often necessary in studies of Mendelian genetics to compare the results of particular crosses with those expected on the basis of a particular hypothesis. For example, if a homozygous dominant is crossed with a homozygous recessive and the progeny self-crossed, then we expect a 3:1 ratio of dominant: recessive phenotypes in the F2 generation (see Chapter 3, section 3.2). This ratio is dependent on the random segregation of homologous chromatids during meiosis. Being a chance event, the ratio will not be exactly 3:1 every time such a cross is performed. It is rather like tossing a coin ten times; on average we expect it to fall five times on heads and five times on tails. However, this will not be the case each time 10 tosses are carried out. We therefore need a method of assessing whether the results of a particular cross are close enough to fit our hypothesis. The statistical method used is the Chi squared (χ2) test. It enables us to assess whether or not a difference between an expected and observed result is significant.

A brief explanation of the test will now be given (for further details of this and other statistical tests, see Mead & Curnow, 1983) and the test will be applied to two sets of data given in Chapters 3 and 5.