The data gathered can be used to perform a number of operations such as mean, standard deviation, variance, correlation etc. Therefore it is safe to say that real data makes it possible to make quantitative classifications. That is why we can say that real data makes it possible to run statistical analysis.
The research has been carried out on the results of 2008 American Presidential Elections. The Exhibit 1 shows actual results of elections. The tables are divided according to percentage lead of each president according to states. The data in Exhibit 2 shows pre election polls for each candidate. The data in Exhibit 2 two has been divided according to agencies which had delivered results or conduction these pre election polls. Column D in Exhibit 2 reflects leads to each respective president in states of polls. The data presented is real in nature for Exhibit 2. This is because the format is percentages of actual responses received from the public. Exhibit 1 also shows actual historical data as the responses are shows as percentage of total votes received by each presidential candidate.
In column E of exhibit 2 we have prepared another category denoted by numbers. This is a better way to convert real scale to nominal scale and then convert it to percentage to get a solution. The number ‘2’ represents a tie, ‘1’ lead of Obama and ‘0’ lead of McCain. If we calculate the percentage of ‘1’ to the entire population we can calculate how many polls considered Obama to win the elections. The percentage of polls that showed Obama as the winner were 71% where as only 10% predicted a tie of votes. This shows another quality of nominal data that it has to be converted into percentages to reach an analysis. The presidential elections did show a victory for president Obama which reveals that analysis using real data was successful in predicting election results.
Nominal data can lead to only qualitative