Pizza Chain Dominos - Research Paper Example

What is Statistics? As we learn from the video, “What is Statistics”, using statistics can help us make clearer and more effective decisions. As the pizza chain Dominos came out with a new product, the deep dish pizza, they gathered data and used those statistics in every step of the process. First, Dominos had consumers do taste tests to rate two different pizza recipes. This is called research and development, and is a process that is used over and over in the creation of new products. Next, they used demographic statistics to find out what their target demographic was so they could cater the new product to that group. Finally, after they introduced the pan pizza to the market, they followed up with customer surveys to see how well customers liked the new pizza. Gathering and interpreting consumer data is crucial the success of a business like Dominos. Statistics help them make accurate assessments of their products. Without the data, Dominos would just be guessing about whether their products are a success or a failure. Picturing Distributions In the video, “Picturing Distributions,” we learn that pictures are very powerful tools for conveying information. Graphic data can help us to easily discover patterns and trends. In statistics, these data patterns are called distributions. One important type of distribution graphic is called a histogram. Histograms can be used to show many types of data. The video shows that meteorologists use histograms to predict the time of day (as well as the time of year) when a lightning strike is most likely to occur. It also shows that histograms are often used to figure out how many people in various age groups watch a certain sitcom. Another type of important distribution graphic is called a stem plot. These are similar to histograms but have the added advantage of showing all the original data. Describing Distributions The video, “Describing Distributions,” describes how a distribution can be used to illustrate the problem of women getting paid less money than men for doing the same job. Two types of data, called the mean and the median, can help us accomplish this task. A mean is also known as an average while a median is the number that splits the data into two parts; meaning that half of the numbers are below it and the other half are above it. As we see from the graph shown in the movie, the median of the amount of money men made in 1987 was significantly higher than the amount of money made by women. In addition to means and medians, other displays such as box plots (helps identify the spread of the data), interquartile ranges (shows the spread of data near the median), and standard deviations (shows how much variation there is from the average) can effectively describe distributions and provide more insight on trends and tendencies. Time Series In the video, “Time Series,” we learn how we can identify statistical patterns over time. In a time series, there are cycles and trends. A control chart is a time series that shows whether data in that series varies a lot or is basically in control. If a distribution is in control, it means that most of the data will end up close to the center, or median, of the graph. This type of chart is important in a variety of settings including the fields of health and industry. One thing to keep in mind is that not all patterns one discovers from a time series are necessarily valid. Statisticians often argue that with certain data sets people are just “seeing what they want to see” and making inaccurate assumptions. These kinds of debates are very common when it comes to interpreting data over a series of time. Correlation The “Correlation” video shows how we can find a correlation coefficient between two variables; for example, how a cake’s taste is related to its price. A correlation coefficient is denoted by the letter “r” and is always a number between +1 and -1. The closer the coefficient gets to one, the more positive the correlation is. This means that as one variable increases, so does the other one. On the other hand, the closer the coefficient gets to -1, the more negative the correlation is. This means that as one variable increases the other one decreases. Additionally, squared correlation coefficient, or “r squared”, can tell us how good of a predictor the modeled values of a correlation might be. The Question of Causation The video, “The Question of Causation,” explores the question: how do we know that all relationships are causal? In other words, how do we know that “A” actually causes “B”? We should always keep in mind that just because there is a strong statistical correlation between two things, it doesn’t necessarily mean that one thing causes another. Another hidden factor could actually be responsible for a correlation. Also, a correlation may just be a coincidence. As the video shows, even the correlation between smoking and cancer is difficult to prove statistically. This is because the cause of the cancer cannot be shown through a simple experiment. Blocking and Sampling The video, “Blocking and Sampling” describes some methods we can use to make sure our experiments are more accurate. If your subjects are systematically different in any way, it could affect the validity of the experiment, which is why you want to do something called “blocking.” Blocking means you put the subjects of the experiment into groups based on their similar characteristics. This method can greatly reduce errors in statistical results. You should use blocking in your experiment if there is any systematic difference that will cause a test subject to react differently to a treatment. Another thing that improves the statistical validity of experiments is called “sampling.” Because it is often unrealistic to test every possible result, sampling tests only a small subset of a population. From this subset, we are able to make inferences about an entire population. From the video, we can see an example of sampling at a potato chip factory. Here, a random chip is sampled frequently to make sure that nothing is going wrong during the production process. Sampling and Surveys In the video, “Samples and Surveys,” we learn how to design an accurate survey. To conduct an accurate survey you should do it in stages. First, you would take a random county and select a sample from that county. Next, you would take a city in that county and select a sample from that city. Lastly, you would take a sample from individual households within that city. You would need to do this process for a variety of similar counties, cities, and individuals in order to get a random sampling. This process is called stratification. One of the top survey companies in the country carefully selects 1500 individuals each year for their survey questions. From this sampling they can get a general idea of the attitudes, opinions, and beliefs of the entire population of the country. As long as all Americans have an equal random chance at being in the sample, the sample should end up being a fairly accurate representation of the entire population. Random Variables The video, “Random Variables,” illustrates that probability often operates on independent events. This means that results that have been obtained in the past don’t influence the future probability because each event is independent of the other. When it comes to random variables, one very significant rule is called the multiplication rule. This rule says that if two variables are independent, then the probability that both events will occur is the product of their individual probabilities. One way we can see how the multiplication rule works is by looking at the probability that a coin flip will turn out to be heads. The probability of this is obviously ½. Also, the probability of getting heads on a second flip is ½ as well. Therefore, the probability of getting heads twice in a row is ½ multiplied by ½, which is ¼. The multiplication rule is very important in a variety of fields, especially in a field where an equipment malfunction can lead to death. That is why it’s important to get the probabilities of equipment success as close to 100% as possible. Case Study In the “Case Study” video, we learn what a powerful tool statistics can be when using all of its various elements together. Without statistics, for example, we wouldn’t know if the prescription or over the counter drugs we take are safe and effective. Inference concepts such as sampling and distribution are extremely effective ways to come to solid conclusions about a data set. However, statisticians who look at the same data often come to different conclusions. However, good critical thinking along with sound research methods and inference techniques can greatly improve the accuracy of your results. Read More