PSY 550 Midterm Exam 2 PSY 550 Midterm Answers

PSY 550 Midterm Exam 2 PSY 550 Midterm Answers Compare and contrast the following terms: (a) test-retest reliability with inter-rater reliability, (b) content validity with both predictive validity and construct validity, and (c) internal validity with external validity. Test-retest reliability and Inter-rater reliability: Test-retest reliability measures the reliability of a test overtime and measures its consistency. For example, if we gave the same test twice to the same people, but at a different time, we could see if the scores are the same. We could test someone on a Tuesday and give the same test again on a Thursday. From there, we could correlate the test scores to one another. Inter-rater reliability is when we are looking at similarities in rating behaviors. This is when two or more raters view the same exact behavior and are in agreement with what the viewed. Both test-retest reliability and inter-rater reliability are forms of reliability. They both measure how stable a measure is over a period of time. Content validity and both Predictive validity and Construct validity: Content validity is when we determine is the correct item or test was used. For example, we are looking at if the test properly measures what we intended it to measure. This is different than both construct validity and predictive validity. Construct validity is whether the scale that was used in the test measures the construct correctly. While predictive validity is determining which scores obtained by a certain measurement predict the outcome that it should be predicting. Content validity, predictive validity, and construct validity are all measuring the validity of a particular measurement. Internal validity and External validity: Internal validity is the extent that we can manipulate a variable and cause a change to a dependent measure. If we have more control over a study, we have more internal validity. Internal validity also refers to the extent of where the experiment or study is free from errors. If any differences occur within the experiments, they are due to the independent variable and nothing else within the study. External validity is when an experiment or study can be related back to the world. External validity is applying the results outside the context of the study. It is when we can generalize results of the study to other people, times, and/or situations. Use Stevens’ model of measurement scales to answer these questions: 1. What type of scale is used when the variable is social class? Explain. 2. What type of scale is used when the variable studied is a Top 40 list of popular songs? Explain. 3. What type of scale is used when the variable studied is intelligence scores (for example, on the Stanford-Binet-R)? Explain. 4. What type of scale is used when the variables studied are low and normal birth weights of newborns? Explain. The scale used to explain social class would be nominal. When talking about social classes, the different classes are categorized, "upper class", "middle class", "lower middle class", and "working class". While they are categorized, the categories do not imply that one is better than the other. The scale to explain the top 40 list of popular songs would be on the ordinal scale. This is because the songs are being ranked in order and a number is being assigned to it. By looking at this data, we would be able to see that #32 is higher than #45. For studying intelligence scores, we would use an interval scale. This is because there are equal differences among the numbers on the scale. They represent equal differences in the intelligence score. In this scale, there is no fixed zero point. When variables studied are low and normal birth rates of newborns, we would use the ratio scale. On this scale, there is a fixed zero point. Weight is one of the few variables that does have a zero point. Define and describe the six steps of the scientific method. Give an example of each step. The six steps of the scientific method: 1. Identify a problem: To first start the scientific method, we have to choose what our problem is and what to investigate. In this step we start to think about how to get answers to what we are trying to find. Typically, in psychology, we are trying to investigate problems related to human or animal behavior. 2. Develop a research plan: Once we understand what we are going to study and we have a hypothesis, we can start to develop a research plan and how we are going to test our variables. In this step, we also need to identify our participants, evaluate ethics and gain approval, and select a research strategy. 3. Conduct the study: Using the plan that we developed in the previous plan, we put it into action. While the research is being conducted, data is being measured and recorded. 4. Analyze and evaluate the data: This is where we are able to review our data and understand what it means to our hypothesis. We summarize the results and data and analyze to give it value to our research. 5. Communicate the results: This is when we show our results in a form of communication. This could be in a written study, an oral presentation, or in a poster. This is also when APA guidelines are followed to help with formatting. 6. Generate more new ideas: This is the last step in the scientific process. Here, we can expand our ideas if our results supported the hypothesis. If the results did not support our hypothesis, we are able to start the project over or reformulate a new research study. Contrast experimental with quasi-experimental methods; generally speaking, which methods are most useful? There are many differences between true experiments and quasi-experiments. In a true experiment, participants are randomly assigned to a group. This group is either a control groups or a treatment group. In a quasi-experiment, participants are not randomly assigned. Also, in a quasi-experiment, control and treatment participants may differ in other ways. Because of this, the researchers that are conducting the experiments have to try to statistically control for many different differences. The experimental method is preferred over quasi-experiments due to the lack of control that quasi-experiments have. The experimental method is more reliable and can be replicated. Researcher Sandy Beach found a correlation coefficient of +2.3 between variables X and Y. She is thrilled because that correlation is “like, way high.” She also feels that this answer “proves that X is like, causing Y.” What are three problems with Sandy’s conclusions? There are many problems with Sandy's conclusion. The first problem with Sandy's conclusion is that correlation coefficients are between -1 and +1. Also, correlation does not prove causation. Just because the correlate with one another, does not mean that X is causing Y. Correlations are also not "like way high". Correlations are either positive or negative. Finally, she "feels" that X is causing Y. The relationship between the variables is not based on what Sandy "feels" but what the data shows. List the main sections (and subsections) of a research report. The main sections (and subsections) or a research report: Title Page Abstract Introduction Opening Literature Review Closing