
If three genes have an order abc, and the recombination values (ab) and (bc) are independent of each other, then one should be able to predict the double crossover value between a and c as a product of individual values (ab) and (bc).
For instance, if (ab) = 10% and (bc) = 20%, the double crossovers in (ac) region would be 10% of 20% = 2%. However, in practice the value of double crossovers is always less than the expected value. This is explained by the phenomenon of interference, which means that crossing over in one region interferes with the crossing over in the adjacent regions. The degree of interference may vary in different regions. If double crossovers are absent altogether, we would say that interference is 100%, while if it equals the expected value, we would say that there is no interference. Another term coincidence is used to express the same phenomenon. 
If double crossovers equal the expected value, there is 100% coincidence, while if no double crossovers are found, coincidence is zero. The coefficient of coincidence can be calculated as follows:
In the example referred to above, expected double crossovers are 0.1 (10%) x 0.2 (20%) = .02 (2%). If the observed value is 1%, coefficient of coincidences, will be 0.5 or 50% coincidence. The greater the coincidence, lesser will be the interference and vice versa.
Contrary to the common occurrence of
positive interference in flowering plants, in bacteriophages and the fungus
Aspergillus, negative interference was observed. It means that recombination in a particular region, enhances rather than decreasing recombination in the adjoining regions. This is partly due to a different method of reproduction in these organisms. It is postulated that in these organisms
effective pairing sties with high recombination frequencies exist.