HOD SYS final exam study guide PDF

Preview

Summary

Answers to all questions for practice final exam...

Description

HOD2500 Final Exam Study Guide

What will be covered? 1. Lecture content 2. Any question from any quiz (even if it is not listed in the Concepts or Big Ideas sections below.) How will you be tested? There will be 200 points possible and your score will account for 20% of your total grade in the class. 1. Short answer (full paragraph) ==> 5 questions @ 10 points apiece 2. Major questions (3/4 to full page) ==> 5 questions @ 30 points apiece Short answer- about half a page to answer- don’t spend more than that Major questions- (3/4 to full page)

Concepts (be able to define, compare and contrast, and give examples of each) 1. Theory and hypothesis -Theory: a means for simplifying and understanding complex realities; it is a model we use to explain what we observe; all models are wrong because it is a simplified version of the world but some are useful; theory of gravity - Hypothesis: an educated guess, often educated by the literature, it’s a clear and directional statement- for example our football team will beat old miss this weekend- it will be tested- it may be based on a theory; and specific prediction of what will happen when we observe something *theorys and models relate to each other because theory is this model we use to explain what we observe and from that model we can generate very specific hypothesis and to the extent those hypothesis seem to be true, we provide support for the theory and false- we need to modify our theory 2. Reliability and validity (including how they relate to each other and how they apply to realworld studies) Reliability: consistency of measurement a. Factors that influence reliability: i. What you measure (physical-easiest to recognize, cognitive- 2nd easiest to correlate, and affective traits) ii. The instrument you use iii. Your technique b. Example: weighing your body and stepping on a scale multiple times and getting very different weights every time, it is not as reliable as a scale on a doctors office Validity: how meaningful, useful, and appropriate our conclusions are The use of the results of a test, not the tests itself 1

c. Example: you go to the DMV and check your eyesight by covering one eye and then the other would be a good test to see if you need glasses to drive, they say yes your vision is good and you don’t need glasses to drive. This is an appropriate test and you can draw valid conclusions from this. but using the eye test is not good to use to see if you can get into Vanderbilt; the test is not valid in determining whether or not you should get into Vanderbilt Related: reliability is a necessary but insufficient condition for validity (must have reliability first and for there to be validity for an instrument)- if you don’t have reliability then you cant make any valid conclusions.

If you think your short on time- first answer question, then finish test and come back and give examples

3. Internal and external validity a. Internal validity: (necessary but insufficient condition for external validity) i. Has to do with our confidence that it is truly the manipulation of that independent variable that made a difference in the dependent variable, and not some other factor ii. Example: whether water or Gatorade produces better performance in athletes; could look at results and say it is Gatorade but someone argues it is external factors that make a difference; IF YOU HAD A CAREFULLY conducted a study and were careful to account for outside variables- a good study- then you would have high internal validity but if you didn’t control for something like altitude- then you would have low internal validity because one game they drink water and there’s low altitude and when they drink Gatorade there’s high altitude. b. External validity: i. Extent to which we feel confident that we can generalize the results from one study to other subjects, settings, and time ii. Argument that is reasonably generalized iii. Example: I’m doing an experiment among test anxiety in Chinese students- and try and generalize it to elementary school students in the usprivate vs public school, all girls vs co-ed school. that has low external validity; different subjects 4. Measurement scales: nominal, ordinal, interval, and ratio (including stats and graphs for each) a. Nominal- named categories; no category is better than another- gender: male/female; religions: jewish/Christian b. Ordinal- also categories but hierarchically arranged- first grade, second grade, third grade- appear in order; first place, second place, second place in the Olympics; socioeconomic status ^categorical data- measures of central tendency the mode; in terms of graphing bar graphs or pie charts -continuous data- histogram/line graphs; median (not sensitive to outliers) and mean; could also do the mode 2

c. interval- equal intervals on some sort of a scale- on a ruler you can go by 8ths of an inch, temperature you can go by degreesd. ratio scale- has a zero on it. Kelvin scale- height- 0 inches tall is the lowest you can gothe benefit is that you can make ratio comparisons- twice as tall equal intervals on some sort of a scale- is ratio this also? *this is particularly important when you graph data- if your graphing ratio data then you screw up the data if you don’t start at zero *measures of central tendency- it doesn’t make sense to say what is the average religion or the average gender… you wouldn’t say average; you would say MODE 5. z-scores and effect sizes (be able to provide equations, calculate values, and explain their uses) a. NOT ON TEST 6. Correlation coefficient (including estimating the size and value…and the effect of outliers) A numerical measure or measure of the strength of the relationship between two variables Look up these drawings A draw an x/y axs and draw a bunch of dots on the chart and look at the correlation .75- looks like corn on the confident estimating size- Tells me how confident I can be that I can guess one variable given the other. If the correlation coefficient goes down to zero If it goes up to 1- you can make a perfect prediction, you give me one variable, I can give you other Estimating value- Positive- one value goes up then the other goes up- studying for a testmore you study higher your scores are Negative- one goes up and the other goes down- example: golf-the more hours I spend playing golf, the lower my score will be Outliers- tend to make the clouds a little larger and make the correlation coefficient closer to 0 because the cloud gets more circular shaped

7. History and role of the Institutional Review Board including autonomy, beneficence, and justice a. Conducted by the military, Japan, and in the U.S. (injected people with drugs to see how long until they die; women who come in for birth control and aren’t given birth control but a placebo and see if they get pregnant) etc… as a result there has been codes conducted b. IRB- a group of people who review proposed research projects and either approve or disapprove depending on how they are ethical standards; it is an approach to make sure our research has guiding principles- IRB does not control what happens at places that do not get federal funding; monitor scientific research at facilities 3

that are funded by the federal government c. Autonomy- respect for persons/animals and what strategies can we use? Informed consent form, voluntary participation d. Beneficence- maximizing benefits and minimizing risks- creating good research designs that you conduct WELL- if you conduct a crappy study- there is no benefit- only a negative side; better be some benefit from your research e. Justice- the goal is to distribute the benefits and burdens of research equally; benefits- make sure people who can receive the burdens are apart of your research-include everyone In your study- you cant just use young, healthy, men and women--- down the road- a pregnant person uses the drug and baby dies- you would have learned that in a clinical trial, distribute the benefits in a just fashionyou go down to a small village and you find out this drug works- people live better for 6 years and you can demonstrate that and you find the results and leave the country- and everyone in the village is like where did the people from the clinical trial go- that’s unjust- if people participate one could argue that you need to continue to receive the benefits 8. Ethnography, case study, phenomenology, and grounded theory a. Book covers (4) and will say which qualitative is used b. Case study- the study of a bounded system, Biography of Steve jobs- case studybounded systems- particular place, person time, event, a specific individual, if you studied Woodstock, or looked at Vanderbilt’s 2012 womens lacrosse season c. Ethnography- the study of culture, people who have shared behaviors and beliefs, about hippies, people in greek orgs, or the mafia of japan; about the culture d. Phenomenology- study of phenomena that transcends culture; ex: slavery, corruption e. Grounded theory- observe the world and generate a theory: i. 5 languages of love- through gifts, things of service- something came up with this theory- Maslow hierarchy of needs- overserving people, interviewing and analyzing and coming up with something and here is the criteria- if there’s a number in the title- 7 habits of this highly effective people, 9 strategies of this etc f. Its this study and here is why 9. Question types: rating scale, rank order, semantic differential, Likert Scale, multiple choice a. Rating scale- rate on a scale of 1-10 b. Rank order- rate these in order these 4 ice creams- rank a bunch from high to lo c. Semantic differential- take two words that have different meaning and put them on two diferent sides- one side “quick” one side “slow”; 2 adjectives that have polar opposite meanings and you decide which one you are closer to d. Likert scale- strong directional statement followed by these 5 statements i. Strongly disagree ii. Disagree iii. Neither disagree or agree iv. Agree v. Strongly agree Anything else is a modified likert scale *You have been chosen by your organization to develop a questionariire- create a statement 4

using the likert and make a way to improve it; add a category such as “n/a or no basis for comparison”, you want to strip out people who don’t have an opinion, or change to completely agree

e. Multiple choice- choose among one of two, or four options abcd

10. Push polling (define and give an example) -Used by politicians where they ask questions under the guides of scientific research as though they are asking questions as if they are trying to learn things, but really thy are trying to change your mind about a topic; influence under the guides of the questionairre Ex: You know chancellor Zeppos has a review of going over budget, knowing this, do you support him building a new stadium? -They don’t ask you about any demographic info that any credible researcher would ask you about -a sneaky technique people use to change your mind under the guides of scientific research -claims are often false or inflammatory, then they ask you a question, take note of your question but they aren’t realy interested in what youre saying but its more a method of influence… if they were doing research they would ask you things such as ethnicity, statistics about you and ish 11. Observer effects and observer bias a. Observer effects- change their behavior as a result of being observed , me influencing people because I’m observing them and they know im observing them, recognizing that people will modify behavior when your observedHawthorne effect; you want to minimize observer effects b. Observer bias- when you get invalid conclusions that are influenced by the researchers perspective ex: for example all researchers have personal biases, if you go into something with a hypothesis, you are more likely to go into something looking to confirm your hypothesis I am a white middle class male so I look through the world through a lenses; drawing invalid conclusions due to biases I carry with me Big Ideas- what you really need to take away 1. What are the ways of knowing we discussed in class? Why are some ways better than others? How you know what isn’t so, you know things based upon tradition or experience or logic or common sense, we know things based upon authority These things aren’t necessarily wrong, but its not necessarily true just because authority said it, you would ask the authority “How do you know?” and if the authority says well based on my personal experience- that’s not very reliable But if they use systematic inquiry (maximize reliability of my data collection, minimize biases, controlled for extraneous variables, carefully analyze results)- this isn’t always reliable, not all scientific research is excellent- a lot of it is poor, its not the perfect way of knowing but it is more reliable than the other ways of knowing 

Personal experience

5













o First-hand knowledge; people can have different perceptions of the same situation o Things often cannot be learned through experience o Seeing is believing, believing is seeing Intuition and reasoning o Common sense o What seems to be true as determined by informal (collective) or formal reasoning o Just because many (or most) people believe something, doesn’t make it true o Conclusions are only as sound as the underlying premises o Example: when someone says something is true and it makes sense, you can likely reason it would be true (think of the obviousness article) Authority o Seeking answers from someone who has had experience with the problem or has some claim to expertise o Authorities can be wrong; they have no claim to infallibility o Authorities may be in disagreement among themselves o Teachers, religious leaders, etc. Traditions o Many people find answers to their questions through long-established customs o Traditions are often based on an idealized past o Traditions can be distant from current realities and the complexities associated with them o Example: customs are effective at some times, but not at others Systematic inquiry o The systematic process of collecting and logically analyzing information to provide explanations o The systematic, testable, and objective nature of research permits the careful examination of the process itself, as well as the results Ex:what are strategies universities can use to reduce drinking underage? Everyone has an opinion based on their background/beliefs; you look at the research and evidence; whose opinion should count for more? Person who has done the research because it is more credible; just because someone had a personal experience does not mean they are right, could be bias How he may present the question: “You’re going home over thanksgiving and have a convo about the strategies a university could use to show how to avoid binge drinking in students. Your sister thinks xyz; your father thinks xxx and represents authority, someone else represents reason.. you come in and show the research... Whose vote counts for the most?” The person who has done the homework aka you; their vote counts for more…you should be able to pick out what each person represents and whose vote counts for the most

2. What’s the risk in believing something simply because it’s obvious? Antithetical People have a tendency to regard any reasonable statement about human behavior as obvious… especially if you provide an explanation. If you take that exact same statement and flip it and provide an explanation, another group will say ya that obvious Just because information seems obvious it does not mean information is true or not *need to be careful about things in class, because just because it’s a reasonable statement so you need to say how is that we know? Once you see studies that prove that then you can know that its true 

The feeling that a research result is obvious is untrustworthy as people tend to regard almost any reasonable statement as obvious

6



When shown results that are reasonable, people think the results are obvious and that they would have predicted the finding, even if it is not exactly true; must look at the evidence to know if something is true

3. What’s the difference between correlation and causality? What’s the benefit of correlational data?  What are correlations: a number that represents the strength of the relationship between two variables (+1, -.2, etc.) (predictive power: the higher the correlation, the more confident you can predict one variable given another; the ability to predict one thing given another); useful in predicting one variable given another but you wouldn’t say one causes another  Correlational data is easy; you just look it up 

To prove causality, must perform an experiment o Experiment: manipulate one variable to see the effect on another o Argue one causes another o Example: spending more time studying, causes you to get a higher score but studying for an exam actually cause your grades to go up? o Best way to find out is to do make an experiment; if there’s a statistical difference then yes- need a good test for causal claims



Problem: when people are given correlational data, jump to a cause and effect situation; people create laws that require/restrict laws based on correlational data o Leads to really bad ideas; no evidence to support ideas How he wants us to answer the question: o “Correlations are very useful because it tells us about the strength of the relationship between two variables. It allows us to predicts one variable given another. But every variable has some type of correlation. Some are high, which actually allows u to pretty accurately to predict the other variable. o If you want to provide strong evidence for cause and effect relationships, you need an experiment. Experiments are very difficult, lots of time, money, controlling for other extraneous variable. o People think that when they have evidence of a correlation it means they have the cause and effect relation. And this is not true (ex. Gun law) . People jump to a cause and effect relationship with just a correlation. The problem is, when people try to jump to those conclusions they then try to create policy (laws etc.) and restrict human behavior. They provide scientific info that is simply false.”



o 



***Why do people still think correlation is equal to causation even if they’ve been taught otherwise?*** o Confirmation bias  If you want something to be true you end up believing it to be true  Stop gathering information when evidence already presented confirms your beliefs Stephen J. Gould o “The invalid assumption that correlation implies causation is probably among the two or three most serious and common errors of human reasoning.” 7

o

***What do you think he means?***  This will be a one page response  It’s all about predicting one variable given another  Correlations don’t prove cause and effect relationship  Experimentation is the best way to prove  Also for sure include discussion about policy decisions… that’s what he said in lecture he wants from us for this question  Kind of repeats question said above just don’t want anyone to be thrown by Gould :)

4. What are major similarities and differences (5 ea.) between qualitative and quantitative research? Quantitative a. Variables? b. Relationship between them? (experimental (causal), correlational, descriptive) c. What are the elements? d. Example: What is the effect of returning home from war on the mental health of United States soldiers? i. What is the ____ for ___? = descriptive quantitative ii. What is the relationship between ____ and ___ for ___? = correlational quantitative iii. What is the effect of _____ on the _____ among _____? = experimental (causal) quantitative 5. Qualitative a. What is the central phenomenon? b. Who are the participants? c. What is the setting of the site? d. Example: How is a woman’s choice to take the full time off for maternity leave perceived by male co-workers in F...