Sampling According to known probabilities, the selection

Sampling is often used statistical tool in our daily life. For example, while purchasing food grains, we tend to check the quality by taking a handful of grains from the bag through visual inspection. It has found its applications in engineering, medical, planning, education, food production, manufacturing industry and so on.

Hence, most of our investigations are based on sample. Sometimes it is possible and practical to examine every person or item in the population. Sampling is considered effective when it is not possible to measure the desired attribute of each person or an item. In all the cases, we believe that the samples give a correct estimate about the population and exact representation of the population. On an assumption that sample possess all features of the population, sampling is turned to be an effective decision making tool. In the situations when complete enumeration of very large statistical data seems to be practically impossible, results are required in short term, area of survey is very large, availability of trained persons and money is limited, sampling is inevitable.    Depending upon the nature of investigation, sampling procedure are classified as:1. Probability sampling.

Don't waste your time
on finding examples

We can write the essay sample you need

2. Non-probability sampling.3. Mixed sampling.

 According to known probabilities, the selection of units from the population is made and this is called as probability sampling. In case of opinion surveys, discretion is used to select the sample units from the population. It is called non-probability sampling. Further, when samples are selected partly according to probability and partly according to a fixed sampling rule, the technique is known as mixed sampling. After all the procedure made to arrive at the conclusion/results for the given population, it sometimes tends to give misleading or wrong information due to personal bias and sampling error. Hence precision and accuracy of the sample with the true value plays an important role to arrive at best results.

It is frequent that precision in obtaining the data of sample is misinterpreted as accurate data, which is not true. This has lead to numerous wrong conclusions and also bad decision making in business, medical and other sectors. In testing and measurement, faulty measurement can be expressed precisely but it must kept in mind that the values are only precise but accurate. For better conclusions, the measurements should be both accurate and precise. Consider the temperature measurement using thermometer given the true value is 260C. On recording repeated measurements if the thermometer shows on an average of 260C +/- 0.10C, then the thermometer is accurate.

If the readings comes to be 26, 26.1, 26, 25.9 and 260C, the thermometer is both precise and accurate.Let us consider a case where the readings be 26.8, 26.7, 26.

8, 26.8, 26.80C, the thermometer is precise to 0.10C but not accurate because the recordings are deviating to considerable scale from the actual value. In short, a measurements can be precise if we obtain similar results with repeated measurement or surveys and a measurement is accurate if we obtain results which are close to the true or actual value.Hence one must be careful and should know precision and accuracy which otherwise leads to making of inappropriate decisions from invalid results about the programmes.

This might further leads to failure in providing needed service or useless resource in providing unneeded service. This indeed a waste of resources, time and money invested as it does not reflect the true situation of the population.In order to get more precision, sample size is increased which may not guarantee the absence of bias producing wrong results and further people may leap an inappropriate faith in inaccurate estimated results. Also quality control is difficult with large sample size. Therefore, it is advisable to go with smaller sample size with less precision, but much less bias.

Because would lead to wrong conclusions, while having less precision may only lead to decrease in confidence level on survey results.


I'm Owen!

Would you like to get a custom essay? How about receiving a customized one?

Check it out