MDM_bias_sampling2

=__Sampling Bias__=

Sampling Bias is the incorrect collection of a sample from a survey in which the sample becomes non representative of the true distribution due to non-random reasons. In unbiased sampling, the differences between the random sample taken and the entire population they represent should be a result only from chance. If this is not the case, and the differences occur not only from chance but also for exterior reasons, then the sample is considered bias.
 * __What is Sampling Bias? __**
 * __

Example of Sampling Bias __** If a poll is conducted on the thoughts of 2000 white middle class university students about their voting intentions for an upcoming election, then the survey will be bias to that group of people. This results in leaving many important parts of the collective whole out such as ethnic minorities, senior citizens, or blue-collar workers. These groups are underrepresented in the sample, making it quite difficult to predict the outcome of the election.
 * __

Sources of Sampling Bias __** · Surveys conducted in certain social class areas due to convenience, which can result in a bias of the area only represents a certain group in the population  · Non-responses cause sampling bias when the reason for the non-response is related to the phenomenon of the study. It is very likely that that certain group of population that does not give a response will be underrepresented.  · Failure to comply to sampling procedures  · Misuse or miscalculation when using measuring devices such as measuring tape. This source of sampling bias usually occurs for surveys that require physical measurements such as height or weight  · Violations of equal distribution such as duplicate listings of persons

Solutions to Sampling Bias __** To reduce sampling bias, it is good to avoid judgment or convenience sampling. This means not to conduct the survey in a certain area based on the fact that it is convenient for the surveyors. Also, it is helpful to make sure the sampling frame matches the target population as much as possible. A sampling frame, which is commonly used in surveys, can make the distribution more even by obtaining background information on the key characteristics of the target population for the sample. Lastly, it is important to simply make sure that the target population is properly defined.
 * __<span style="font-family: 'Arial','sans-serif'; font-size: 10pt; line-height: 115%;">


 * __Types of Sampling Bias__**

<span style="font-family: 'Times New Roman',Times,serif; font-size: 10pt; msofareastfontfamily: 'Times New Roman'mso-ansi-language;">**__Self Selection__** Self selection bias is the tester making it possible for the surveyor to not participate during the whole survey or certain parts of the survey. If the tester makes a question that that a person may not feel comfortable answering because it is embarrassing to reveal the result of there answer then they can just simply not answer the question. This creates a bias because you can’t get the proper results if people who don’t feel comfortable answering a question just don’t answer it. For example if the question is how often do you participate in illegal activity if the person chooses not to answer they could have been one of the people who would’ve answered yes and without getting the people who would’ve said yes there would be a bias that most people never participate in illegal activity when it may in truth be even or more people participate in illegal activity than those who don’t.

Pre screening bias is setting up your survey so that you get certain types of people for certain types of the survey making your results inaccurate. An example of pre screening is surveying people on good ways to diet at a fast food restaurant. Generally people who eat fast food don’t have the best ideas or ways to diet meaning that you will probably get the biased results of people that don't generally eat healthy and therefore don’t know how to diet very well. Another example would be if you posted flyers to participate in a survey on what your favorite television show is outside of a all women’s gym you will most likely get only women to participate and you would have to make sure you said only women participated with your ending results, you wouldn’t be able to use those results as a general population of both men and women. Selection Bias __** A selection bias is when you don’t include all the categories of people you need to give accurate results for your survey. An example of this would be if you put a sports survey outside a hockey arena you would only get the result of hockey players not soccer players football players and all other sports this would give you a selection bias. Fixing Sampling Bias Statistics __** To fix a sampling bias without changing the survey or if you realize you have done a sampling bias after your survey has already been completed first you have to make sure that entire segments of population have not been excluded from your survey. Once you have done this you would have to calculate the bias for example if you had 500 men and 250 women fill out your survey. Then you would have to calculate the difference in this case its 2 – 1 (500-250) so for each vote of the men it would be worth 0.5 in the results and the women vote would still be worth 1 per vote.
 * __ Pre Screening __**
 * __
 * __

Questions

1. Nike is doing a survey on the comfort of there shoes in comparison to other companies. They want to make sure that they get the right statiscal results of the overall population of the consumer. 45% of nikes shoe consumers are women the other 55% are men but for there survey they get the result of 1000 men and 1200 women how would you fix this sampling bias to get the right statistics to consumer rate?

2. Nike wants to repeat there survey getting no statiscal bias how do they make sure that this happens give three tips on how to get the right number of people in comparison to the population of the consumer to make sure there is no statistical bias

3. Addidas See's that nike is doing a survey for there shoes so they decide to do the survey for there consumers. They give there testers the option to answer certain questions. Will this give accurate results if not what type of sampling bias is addidas using?

//Bibliography Apa Format Sampling Bias//. (n.d.). Retrieved October 28, 2009, from [] > || Remove-item-icon || > || Edit-item-icon || //Sources of Bias in Survey Sampling//. (n.d.). Retrieved October 28, 2009, from []
 * || [[image:prestegeworldwidesamplingbias:/images/remove-item-icon.png?1235627343 width="15" height="16" caption="Remove-item-icon" link="prestegeworldwidesamplingbias:home#"]] ||
 * [[image:prestegeworldwidesamplingbias:/images/edit-item-icon.png?1235627343 width="16" height="16" caption="Edit-item-icon" link="prestegeworldwidesamplingbias:home#"]] ||
 * Discipline. (n.d.). //JSTOR: An Error Occurred Setting Your User Cookie//. Retrieved October 28, 2009, from []