Business Research Into the Legalization of Marijuana

Business Research

Research into the legalization of marijuana will seek to determine what people's views on the issue are. The population is going to be the entire United States, as we are seeking to determine national views about the issue. The sampling will be done at random, using databases of phone numbers and computer-generated dialing. The operators will only know of and gather non-personal information such as age bracket, gender and state of residence. The major downside to this form of sampling is that many younger people will not be sampled because they often do not have landlines.

Random sampling is the "purest form" of probability sampling in that it is free from researcher bias and is more likely to accurately reflect the total population as a result (StatPac, 2012). Landline ownership is a reasonable proxy for the total population, with the caveat noted above. However, other sampling methods are less desirable for other reasons. Systematic sampling with a large population is almost as effective as random sampling. Stratified sampling would be useful if the number of respondents was going to be lower. The need to stratify the sampling by age, gender or location is removed when the sample size is large enough, as this survey is expected to be. A less serious study of this issue would be conducted using convenience sampling, such as asking people at the mall. However, in order to determine from a business perspective what the attitudes of Americans are with respect to the legalization of marijuana, a large sample size is necessary and therefore random sampling is not only possible but the optimal sampling method for the study.

A sampling frame is the actual set of units from which a sample has been drawn (Statistics.com, 2012). The population is adults in the United States, but the frame is somewhat smaller. The first is that the person must have a landline at home. The second is that this landline number must be available in a public directory or purchased directory. The third is that the person is not on the national do not call list. In addition, since the survey is going to be conducted during business hours, the person is going to be likely to be home from 9am to 5pm on Monday to Friday in order to be included in this survey. In addition, this is going to be the type of person who generally would answer a telephone survey, and has 10-15 minutes to do it. Thus, the sample frame is a much narrower segment of the population. However, this frame is acceptable for the purposes of this study if only for pragmatic reasons.

The appropriate sample size for a study of this type must be large enough to reflect accurately the entire nation and to deliver a statistically-significant result. The sample size is therefore determined by the size of the population, the confidence level and the confidence interval. For the entire United States with a 95% confidence level and 3% confidence interval, a sample of 1067 will be required (CRS, 2012). This is the sample size that will be used, so we will need to start with a much higher number of phone numbers in the system. The sample size was determined using an online sample size calculator.

The poll results will be expressed as percentages. The questions will be framed in the multiple choice, yes or no, or other methods that will allow for the expression of responses to be in percentages. An example would be "Are you in favor of the legalization of marijuana?" As a yes or no binary choice question. The results will then be expressed as a percentage, with the confidence interval noted.

Ethical Considerations

There are several ethical considerations in this kind of research. One issue is that of informed consent, where the person being surveyed needs to be aware of what the survey is about and agree to do it. This issue will ensure that all participants are willing and that they understand what they are doing and why. There will not be any payment for participating in this research, as that might also introduce biases.