Stats1 - Essay Example

For example, one may talk about the population of all individuals in US whose IQ is over 120; or one may talk about the population of all electric light bulbs ever produced by Phillips. Every population is characterized by measurable quantities called parameters. If parameter of a population is known, the whole population is known completely. Sample is a part of the population. The quantities measured from the sample are known as sample statistics. Sample statistics are used to estimate population parameters.

While in a given population the parameter is constant, the value of statistic varies from one sample to the other. The population mean is a population parameter while average computed from a sample of size 20, say, from the same population, is a sample statistic which may be used to estimate the unknown population parameter. A different sample of size 20 may provide a different value of the sample mean. 3. When a sample is observed instead of the whole population, sampling error is caused. A sample is observed to understand the whole population.

The population parameter is not known, but is estimated using the corresponding sample statistic. The difference between the parameter and the statistic is a measure of the sampling error. 4. An experiment is conducted under the control of an experimenter. An experiment is opposite of an observational research, where the researcher observes the study units and records observations. In an experiment, the researcher sets up the experimental conditions and controls them as per his/her research hypotheses. 5. This is an experimental study.

Correlational study finds correlation among different variables from the same group of observations. Here two equivalent groups are compared in terms of effects of breakfast on performance. Finding correlation is not the aim of the study. But testing whether there exists any difference between the two otherwise equivalent groups when treated with two types of breakfast is the main objective. There is one case and one control group. Hence this is an example of experiment. 6. Nominal scale is essentially a classification.

Even though the classes may have numerical identifications, such as 1, 2 etc, their nomenclature is important but not the values. Suppose in a town there are two groups of hospitals: general and mental. If general hospitals are called Group 1 and mental hospitals are called Group 2, this will be an example of nominal variable. If we interchange the order, there will be no effect. For ordinal scale the ordering is important, but by how much one level is more than the other, cannot be measured.

Suppose blood pressure is classified as normal, high-normal and high. In this classification there is a hierarchy but the difference between normal and high-normal and high-normal and high cannot be quantified. Interval scale can compare the lengths of intervals. For example, if a flight starts at 6:00 hrs and reaches its destination at 8:30 hrs and another flight starts from the same origin at 6:30 hrs and reaches the same destination at 10:00 hrs, the lengths of the flights are comparable and the difference measurable.

For interval scale measurements, the beginning and the end are fixed. In mathematical term, in interval scale there is no concept os an absolute zero. But it contains more information than ordinal scale variables, since its lengths are quantifiable, which is not possible in case of ordinal variable.