The post How to Calculate Sample Size in a Research appeared first on Reading Craze.

]]>You need to know the following before calculating sample size.

At what confidence level you want to test your hypothesis. Suppose you want to test the hypothesis at 95 % confidence level, it means that you are 95 % confident about your findings. The chances of error in your study are 0.5 %.

The population size needs to be known and it is known in majority of the studies if the population is small. In case the population is large the size is unknown. Sample size can be calculated even if the accurate population size is unknown. An approximation of the population can be made in case you do not know the exact number of the people in your study population.

You need to know how much variation there might exist among the values or responses that you will get for your questions.Since the study is still not done you do not know the variation or standard deviation. You can take a standard deviation of 0.5 or 0.1 for example. This is only a probabilistic deviation and not the actual one. A 0.5 deviation means that there will be great variation in the responses and data will be spread over a large range. A 0.1 deviation means that there will be less deviation in the responses and you will get a range of data that is closer to the mean.

Sample mean = X

Population mean = mu

Sample size = n

Standard Error = E

Standard Deviation = s

Critical Value = za/2

The formula for standard error is

By rearranging this equation and adding the value of critical value we can get the formula for sample size.

The post How to Calculate Sample Size in a Research appeared first on Reading Craze.

]]>The post What is Sampling appeared first on Reading Craze.

]]>Take an example you want to know the average age of the students in a class. There are two ways to do this. You either ask each student their age and then sum them up and divide them by the number of students and in this way you get the average age of the students in the class. This method can be very time consuming and at times impossible to do. Another way to do this is to ask few students their age and sum them up and divide them by the number of the students and you get the average age of the students in the class. This method will be easier and convenient and can be used even if the number of students is too large.

Take another example you want to take a poll about a TV programme. You can submit a questionnaire or ask all the viewers who watch that programme to know their opinion. This can be daunting and at times very impractical. The better way can be to ask a few people and generalize their opinion to all the people who watch that programme.

Without sampling you cannot continue some studies, especially where the population is too large and widespread. Suppose you are conducting research on the consumption of electricity in an area, you cannot reach every household to know how much and how they use electricity in their houses. You can take sample and ask them and get results. Hence we can say that sampling makes every research possible.

Your time, energy and money everything is easy to manage when you take a sample rather than studying the whole population. This is important because in most of the studies you have a limited time and you have to find out the results during this time. When your research takes too much time it becomes obsolete till you are able to implement its findings. Its always better to be done on time.

In any way there will be difference between population mean and sample statistics. This is called as sampling error. There are always chances of error and this error should not be high. A small error is expected and 0.05 or 0.02 percent error is thought to be acceptable but there should not be greater error. There are several ways to find out how much confidence we can place on our sample.

Sampling is a trade-off between gain and loss but you cannot take on your study without sampling. The researcher should ensure that the sample is free of biases and it is highly representative of the population.

One of the most important factor in determining the accurateness of the sample statistics is the sample size. Your sample should be enough large to represent the population. In a large population a small sample cannot yield accurate results.

When the population is highly varied the selection of sample becomes difficult. In this case the population is divided into groups and from these groups small subgroups are selected as sample. The more varied the population is the more groups and subgroups are drawn before the final sample is taken.

The post What is Sampling appeared first on Reading Craze.

]]>The post Sampling Terminologies appeared first on Reading Craze.

]]>In an ideal situation the whole population should be taken to study any variable that you want to study. It is unfortunately impossible to study the whole population. In most of the cases population is large and cannot be studied as a whole. A sample is a part of the population that is drawn from the population. It makes the collection of data and its analysis easy and feasible. Sample size is denoted by lowercase letter “n”.

Population is the total entity of people upon which you want to generalize your research. In research population size is denoted by uppercase letter “N”.

Target population is different from the actual population in one way. Target population can be very large and the researcher might select a small group of that population or people living in one area to make the research easy. For example, the researcher wants to study the effect of one beauty product on elderly ladies. The target population in this case will be all ladies that use that product. This can be very large group spread over several states or even countries. He decides to select a population that is accessible or lives in his region. This will be called as sampling population or population. Hence target population can be very large as compared to the population that the researcher selects for generalizing his findings.

Sampling framework is the entire population of people, situation, incidents or households from which the researcher has to take the sample. There might be several sampling framework but it is not always possible to draw samples from them. Some frameworks are difficult to research because of social, moral or ethical issues. Sampling framework is that population from which you can draw sample feasibly.

Sampling design is a technique or a procedure to select samples from the target population. The sampling design ensures that each element in the population has an equal and unbiased chances of becoming part of the sample. Sampling design can be nonrandom and, in this case, some compromises are made to get the desired sample.

Standard error is the standard deviation of the sampling distribution. Standard error thus describes the chances of error that may occur in the statistic calculated from the sample. As sample could just estimate the characteristics of the population, hence standard error describes the possibility of error in the sampling distribution.

Standard deviation determines variability of the actual population from which the sample has been drawn. It is denoted by Greek letter sigma and often written as sd. Standard deviation is also the square root of the variance of the population. In statistical terms it can be defined as the deviation of the data points on each side from the mean.

Confidence interval and confidence level are the two terms that are constantly used in sampling and in analyzing the data. Confidence level is the degree or percentage of confidence that the researcher has on the estimates that he has drawn from the sample. It means that higher the percentage of confidence level, greater will be its reliability.

In statistics mean is the most common type of average that is used. It is used in sampling to describe the characteristics of sampling distribution.

The post Sampling Terminologies appeared first on Reading Craze.

]]>