Questions on Varied Topics
- What is informed consent, and why this is an important concept in modern research studies? Just what is involved in informed consent and how can we do our best to make sure that when we do get consent from a participant to our study that it is actually “informed”? Why even bother? What is your own take on this topic?
An informed consent is the permission sought by a researcher, or the one who is conducting the study from the respondent or the person participating in the research before conducting the study along with informing them about any implications and consequences involved. As such, the individual concerned possesses sufficient reasoning ability to understand such implications, which constitutes an essential part of the informed decision It also means the subjects under study understand the study. Informed consent involvers telling the participant what the study is about and what it hopes to achieve. Getting informed consent means telling the participant what the study entails. The participant can then give the consent with the knowledge of what is to happen. It is important to get an informed consent as this is required by law and needs to confirm with the medical and research ethics apart from maintaining full transparency in the matter. It is mandatory to get an informed consent for both the participant and the individual conducting the study.
- Let’s take a look at correlation!! What is correlation and how does it work? What does “r” stand for, and what is a negative correlation and how would we know it is a negative correlation when we look at the analysis? ………… (One more thing, :-)) let’s try to look around at some literature and find an example of correlation from a report or article we can share.
A correlation can be defined as a relationship between two random variables or two sets of data. It is represented by a single digit which acts as an informant of the extent of a relationship between the two variables. Its utility lies in giving indication of the predictive relationship which can be made use of in daily life. A correlation works by testing the relationship between two variables by calculating the correlation. For a correlation to work, the data needs to be quantifiable. The “r” in this particular case represents the connection that the two particular variables possess. A negative correlation means the computability of the two variants is not pronounced, and an inverse relationship exists between the two. It assumes a value between 0 to -1.00. It illustrates that in a bivariate plot, there would a negative slope, flowing from the left to the right. In a negative correlation, when one variable is higher, the other is lower. An example of a correlation is the relationship existing between height and weight of a person.
- Our text mentions a social artifact……… what the heck is this and how does the “social” part of it fit in?
A social artifact is derived from the word cultural artifact. It illustrates anything created by humans. It makes available information about the creators and the users. Nevertheless, unlike an archeological artifact, a social artifact is not always in a physical form and does not necessarily require possessing historical value. As in other terms of artifact like “cultural”, which entails something that represents the culture of the creator and “archeological”, which illustrates an artifact recovered archeologically, “social” in this case represents something that has social value or something that represents the society. In essence, it represents an item of cultural value to the society, and thus a cultural artifact. Their significance lies in their giving insight into economic development, social structure, and technological process among other attributes.
- If you were instructing this class how would you explain the concept of reliability to students who have never had a course in research or statistics? What do you think about the methods that our classmates share with us?
Reliability in layman’s terms is the reliability of the measures. It depends on the repeatability of measurements, making them reliable. It is intricately linked with the quality of measurements. It measures the consistency of measurements. In terms of research and statistics, reliability is the degree of the consistency and stability of assessments tools. The methods shared in class of reliability are consistent with the requirements of reliability. They illustrate the consistency and repeatability of results. It needs to be understood that reliability cannot be calculated, but only estimated thereby giving rise to different types of reliability.
- Pick one of the types of validity and tell us how it works………by the way…….. your students have no idea what you are talking about, so do take the time to explain it well and in simple terms, perhaps with an example or two. Have fun!! 🙂
The believability and reliability of research is referred to as validity. The significance of validity lies in its ability to help researchers determine the type of tests to be used and use methods that are ethical and cost-effective and truly measure the construct in question. There are several types of validity, but the validity in focus this time is the construct validity. Constructs validity in research defines the measurability of a test or an experiment to its claims. There two forms of construct validity, convergent validity (confirms that constructs are related when they are expected to be related) and discriminant validity (tests constructs with no relationship and confirms their lack of relationship). An example of construct validity is when measuring depression. The test should only measure depression but no other factors like stress or anxiety.
- There are various methods of sampling that may be used depending upon our research question. We are reading quite a bit about these methods and I was wondering why we would want to use probability sampling rather than non-probability sampling and how does our method of sampling affect our level of error? What do you think?
Probability sampling represents any sampling method that selects its samples randomly. Probability sampling uses random selection. On the other hand, random selection is not integrated in non-probability sampling. Using probability sampling we can rely on the rationale of the probability theory. It is possible to estimate or calculate confidence intervals when we use probability sampling. Non probability sampling and probability sampling affect the level of error differently. The level of error in non-probability sampling is higher than the level of error in probability sampling. From the information above, it is clear that using probability sampling is better than utilizing non probability sampling.
- What is the difference between stratified sampling and quota sampling? Is one method more likely than the other to produce better results with respect to a lower rate of error in our sample analysis? Why do you think so?
Quota sampling is a representation of non-probability sampling while stratified sampling represents probability sampling. In stratified sampling, the sample is chosen at random and divided into exhaustive and mutually exclusive groups while in quota sampling the population is divided into mutually exclusive sub-groups then the sample is chosen through judgement. In terms of the best sampling method between quota and stratified to produce the lower rate of error, the stratified sampling method comes out first. Stratified is better due to its definitive sampling. Choosing the samples randomly eliminates any risk of biasness or error that might be associated with quota sampling.
- As the chief of police you recently attended a conference where a new method of patrol was discussed. You are interested in this new method, but hesitate to implement it in your city-wide and then find out that the new method of patrol really does not work that well. How would you test this new patrol method to see if it really does reduce the fear of crime so that you have enough information to make the decision to implement it in all precincts or not? Okay class, kick this one around and see what you can come up with!!!
For the chief of police to understand whether the new method can apply to his station, he must carry out a test trial. He must conduct a survey that will aid in knowing whether the method is reliable. I would test this patrol by selecting a sample of officers from the whole team of officers and direct them to implement the method individually in the way they take their patrols. After that I would record the results and then conduct the analysis that will look to verify whether the new method of patrol is more effective than the old one. Through the results derived from the analysis, I would then decide whether to implement or reject the new method of patrol.
- What the heck is multi-stage cluster sampling? How would we use this method and why in the devil would we want to? How about we mix it up a little more and throw in some stratification, or for even more fun some disproportionate sampling and weighting? Can you think of an example of when we might select to use these techniques…………….and, is there a problem or two we might encounter?
Multi-stage cluster sampling represents the division of larger clusters while sampling into smaller, targeted groups. We would use this method by subdividing the population we are studying into clusters, then divide the clusters into groups and survey them separately. We can use this method to study large populations as it reduces costs, minimizes errors and reduces logistical problems due to the large population size, an example is government surveys. Disproportionate sampling and weighting is the case whenthe sample in question is too big or too small compared to the others but shares the same chance of success. Stratification involves dividing sampling into strata and choosing a sample from each strata.
- There have been a number of experiments performed in the past that have become quite controversial today. In fact, some of these experiments lead to the development of the Internal Review Board requirement that we know today. We learned a lot about human behavior from the Stanford Prison Experiment, and no one really was hurt, so what was the problem with this university-based research project? Why would we not do a similar experiment today for the betterment of mankind?
The Sanford Prison Experiment was wrong because the experiment abused the human rights of the prisoners. As the Experiment was an attempt to investigate the psychological effects of the perceived power, with a focus on struggle between the prisoners and the prison guards, an element of authoritarianism and psychological torture got infused in unwittingly. Also, the experiment was forced on the prisoners even when they refused to continue their participation. The experiment had the potential to have adverse effects on the physiology and psychology of the prisoners. The experiment would not be repeated today due to the ethical guidelines that need to be followed, and due to the emotional and psychological and emotional trauma it is feared to cause. Yet another reason is possibility of development of sadistic tendencies unconsciously, which will be repugnant to the basic purpose of any such experiment.
- What is the difference between open-ended and closed-ended questions, and what are the advantages and disadvantages of both methods? Then, is there a preferred method of using them in our questionnaire and where and why would we use/place the questions that cover the demographics?
Open-ended questions are those which are exploratory in nature and do not offer choices at the end while close-ended questions offer the answerer multiple choices which a person can choose from. Open-ended questions provide rich qualitative data, and therefore take time to comprehend. On the other hand, close-ended questions are conclusive in nature, which require a researcher to have a clear understanding of the subject matter of the questions before framing them. Both types of questions are used differently. Questions that cover the demographics are placed on the close-ended questions segment because such questions are easy to code which makes them particularly useful when trying to prove statistical significance of a survey’s results. Close-ended questions are easier and quicker to answer but can also coerce answers from respondents with no information. Open-ended allow for unlimited responses but can also lead to misleading answers.