What does the margin of error tell us about a sample taken from a large population? How does the confidence level for a sample outcome differ from the sample’s margin of error?
The margin of error is a numerical value that expresses or estimates the likelihood of error in an experiment’s results. Thus, it can somehow account for the statistical significance of a result. If it is smaller, then the likelihood of error is also small, making the results more credible. Smaller sample sizes from a large population generally have a larger margin of error while larger samples taken from the same population have a smaller margin of error. This means that the results that can be obtained from an experiment can be made less erroneous when the sample obtained is larger. A larger sample can more accurately illustrate the behavior of the population and reduce the margin of error in an experiment. For instance, the margin of error is an indication of the closeness of the results obtained from a sample to the results that can be obtained from the whole population. Thus, a smaller margin of error is desired. If the fluctuation in the results obtained from surveys is within their margin of error, then it can be safe to conclude that the results are accurate and that the changes could have come from sampling errors. Deviating results that go over the margin of error indicates that the change is statistically significant and should not be interpreted as simple errors in the sampling (Carey, 2012). Moreover, the confidence level is the probability that samples that is expected to fall under the conclusion or a range of values. For instance, if it had been established that there is a 95% level of confidence that a value will fall in a given interval, then we can expect that 95 % of all values will be conclusive of the hypothesis while the remaining 5% will be off the interval. This 5% is considered irrelevant for some since there are expected fluctuations (Hunter, 2016).
What is a null hypothesis in causal research and what does it mean to say that a study has failed to reject the null hypothesis?
The null hypothesis is the hypothesis that claims that there is no difference between the results of two groups (Carey, 2012). In other words, the two variables in the experiment that a study is trying to compare are not significantly different from each other. This is usually the hypothesis that researchers try to disprove or to reject (Gonzales, n.d.). For example, a researcher would like to determine if giving birth to a son during nights when there is a full moon are is common. The null hypothesis would be that there is no statistically significant increase in the number of sons being born during nights when there is a full moon. The researcher wishes to reject this null hypothesis to claim that when there is a full moon, a woman is more likely to give birth to a son. When a study has failed to reject the null hypothesis, then the study had failed to illustrate that a factor influences the results of the experiment. The influence that a factor has on the samples are not enough to say that it is relevant parameter. In the aforementioned example, the causal factor is the full moon while the result is the increase in the percentage of people giving birth to sons. Failing to reject the null hypothesis in this case means that the percentage of male births during nights with a full moon and nights without a full moon does not indicate a statistically significant increase in the number of sons being born. They may be unequal at some extent, but is not enough to reject the null hypothesis.
How do randomized, prospective and retrospective studies differ from one another? What are the major advantages and disadvantages of each type of study?
Randomized studies or randomized casual studies is a study where samples used for the experiment are taken randomly and assigned into experimental groups or control randomly before exposure to a causal agent. Randomized studies are advantageous since it is able to provide indisputable results that can be used for causal experiments. However, it has several disadvantages. Randomized studies are generally expensive, especially when the experiment is dealing with large samples. It may also take longer to observe results if the samples display the effects of the causal agent indefinitely. Thus, randomized studies are costly and time inefficient, but offer unequivocal results.
Prospective studies, on the other hand, start with an experimental group where samples that have been initially exposed to the causative agent being studied and a control group where samples have not been initially exposed to the casual agent. Observations over time are noted to determine the effects of a causal factor. An advantage of prospective studies is that it is easier to carry out than randomized studies since there is no need to directly expose samples to the causal agent. Thus, it is less expensive and does make issues regarding ethical concerns. It is likewise at a disadvantage since it rarely allows researchers to study larger sample sizes. A prospective study also does not fully take into account the differences between the effects of the causal agent that would require larger sample sizes to fully understand. Lastly, a retrospective study takes a sample with those who were exposed to the causative agent and those who were not. This type of study involves looking back at how differences in the degree of the causal agent affects the samples. Similar to prospective studies, retrospective studies are relatively easier since there is no need to expose samples with the causal agent. Thus, making it also cost efficient. However, this does not provide the extent of difference among the effects (Carey, 2012).
Describe each of the fallacies listed below and make up an example of each
False anomalies
False anomalies means omitting information or facts that would make a result explainable. Thus, making something stranger than it should be or more mysterious. This type of fallacy can often be found in documentaries or books regarding UFOs. Information that can explain how things seem to appear may have been omitted intentionally to make an anomaly that captures the attention of some (Bluedorn, 2004).
Questionable arguments by elimination
Questionable arguments by elimination is a reasoning which involves the acceptance of an information due to the falsity of another. An example of which is if people are provided information that a certain product is produced under certain standards and made to believe that products other than the said products are harmful (Bluedorn, 2004).
Illicit causal interference
This means stating that an agent causes an effect when in fact it is only correlated and there are many other factors at play. An example is when people explicitly state that consuming fatty foods lead to heart diseases when it there are many other factors like exercise at play (Bluedorn, 2004).
Unsupported analogies and similarities
It is a fallacy that focuses on the similarities when making an analogy, but disregarding the differences. An example is when simple predictions are compared to well-tested scientific theories since the make sense,but disregarding their obvious differences (Bluedorn, 2004).
Untestable explanations and predictions
This is a fallacy that offers explanations and predictions that cannot be tested or validated, leaving explanations that are unchallenged. Conspiracy theories often use this fallacy since these theories can rarely be tested in real life (Bluedorn, 2004).
Empty jargon
This is a fallacy that incorporates usually used scientific terms or jargons to appear more valid. It tends to overwhelm people by the scientific jargons, making them stop questioning a notion. An example of which is if alternative medicine are described to target diseases that sound technical or use scientific words to appear effective (Carey, 2012).
References
Bluedorn, N. (2004). A Beginner’s Guide to Scientific Method by Stephen S. Carey. Fallacy Detective. Retrieved July 26, 2016, from http://www.fallacydetective.com/articles/read/a-beginners-guide-to-scientific-method.
Carey, S. S. (2012). A Beginner's Guide to Scientific Method (4th ed.). Boston, MA: Wadsworth, Cengage Learning.
Gonzales, K. (n.d). What is a Null Hypothesis? - Definition & Examples. Study. Retrieved July 26, 2016, from http://study.com/academy/lesson/what-is-a-null-hypothesis-definition-examples.html.
Hunter, P. (2016). Margin of Error and Confidence Levels Made Simple. ISixSigma. Retrieved July 26, 2016, from https://www.isixsigma.com/tools-templates/sampling-data/margin-error-and-confidence-levels-made-simple/.