Mean Hb values are compared in two population groups. The best test is -

Ans. is 'b' i.e., Unpaired T test Student t-testo Small samples do not follow normal distribution as large ones do. To overcome this problem T test was devised by W.S. Cassette whose pen name was 'student' and hence this test was called student't' test.Application of test :o It is applied to find the significance of difference between two means, when sample size is less than 30.o It is applied on a standard population either as :Unpaired 't'test: Applied when data is unpaired, i.e. when separate observations are made on individuals of two separate groups, and these need to be compared :Null hypothesis is assumedObserved diference between means of two samples is found.Standard error or difference between two means is calculated.t value is estimated. This is the ratio between points 2 and 3.Degrees of freedom are next calculatedCalculated values are compared with the 't' table values at the calculated degree of freedom (option d )P values are obtained and if it is more than 0-05, it has no significance.Paired test: Is applied to paired data, when each individual gives a pair of observations such as : when observations are made before and after the play of a factor e.g. pulse rate before and after a drug.Further, it proceeds similar to the unpaired test.Comnwinlv used Statistical testsDataQualitative(Proportions, percentage, fractions)Quantitative(mean & SD. of BP. Sugar, cholesterol etc)Large sample size < 20% of cells have expected value below 5Small sample size > 20%. of ceils have expected values below 5Two sansple group comparisonMore than 2 sample groups Sample is smallSample is large <n > 30) Cbi-square testFisher's exact testStadent 't-test'Z- testANOVA