When assigning subjects to both experimental and control groups, what's the suggested percentage distribution for mentioned groups?
a) 20%
b) 30%
c) 40%
d) 50%
e) 60%
The answer is d) 50%, so both groups may have the same amount of representative subjects.
Place for Section 101 and 114 to post reading memos, practice quiz questions, or study tools.
Thursday, October 28, 2010
Masculinity Quiz
Which of the following statements is not a reason men are more responsive to gender identity threats than women?
A)Masculinity is more valued than femininity in most social contexts.
B)The hegemonic masculine ideal can never actually be completely achieved.
C)Lower testosterone levels lead to an overcompensation.
D)Increasing one's masculinity necessarily diminishes the masculinity of others.
Answer: C. Men with higher testosterone levels are more sensitive to masculinity threats than those with lower testosterone levels.
A)Masculinity is more valued than femininity in most social contexts.
B)The hegemonic masculine ideal can never actually be completely achieved.
C)Lower testosterone levels lead to an overcompensation.
D)Increasing one's masculinity necessarily diminishes the masculinity of others.
Answer: C. Men with higher testosterone levels are more sensitive to masculinity threats than those with lower testosterone levels.
Wednesday, October 27, 2010
Willer's Test of Masculine Overcompensation
1. According to the results of Willer's study on masculine overcompensation, what role did high levels of masculinity play in determining whether or not men would show support for the Iraq War, dislike of homosexuality, and favored SUV's?
A. Attribute
B. Sufficient Cause
C. Variable
D. B and C
E. None of the above
Answer:
Best answer is B. Sufficient Cause
A Sufficient cause represents a condition that if it is present, gaurantees the effect in question. In this case, masculinity could be seen as a sufficient cause to their attitudes on Iraq War, homosexuality and SUV's. If a man carries the presence of (high levels) of masculinity it guarantees their positive views on the war, against homosexuality and would even favor buying an SUV.
2. If masculinity is an independent variable, would threatening that masculinity, effect the dependent variable (which is their attitude on the war) as a moderator?
T or F
Answer: T
Because masculinity itself already shows high levels of support for the war, threatening it would only increase that support.
Overcompensation Quiz Question
Which of the following was NOT one of Willer's findings to his experiment/s?
a) when masculinity is threatened, men express higher masculine attitudes
b) when masculinity is threatened, men take on higher dominance attitudes
c) perceived threats to masculinity only occur at the individual level
d) higher testosterone levels is associated with higher responsiveness to threats to masculinity
c) perceived threats to masculinity only occur at the individual level
All of Willer's findings can be found on page 35 of his study (pg. 146 in the reader). Willer's third experiment actually found that cultural trends (i.e. feminism, homosexuality, etc.) also could be considered threats to masculinity, thus demanding an overcompensation for masculine traits and attitudes.
Quiz Question: Willer - Overdoing Gender
Study 3 (results of a national survey) was able to address which issue of studies 1 and 2 (both laboratory experiments)?
A. Response rate
B. Overcompensation
C. Causality
D. Generalizability
E. Spuriousness
Answer: D. Generalizability
"The greatest weakness of laboratory experiments lies in their artificiality." -Babbie Ch. 8
Because lab settings are so controlled and unlike the real world, it is cautionary to use its findings to make generalizations about the entire population. Furthermore, Studies 1 and 2's participants were all undergraduate students of a large, Eastern private university. Study 3, on the other hand, surveyed people of differing age, gender, race, educational attainment, region, as well as whether they were California residents. This gave the study much more external validity because the research subjects were much more representative of the population.
The answer is not "response rate" because, if anything, national surveys would have greater issues of response rate than lab experiments. Those who participate in lab experiments are often individuals who choose to participate for one reason or another (e.g. money or extra credit) so therefore would respond during experimentation. Those who receive surveys, however, do not provide consent before receiving surveys so would be less likely to respond.
The answer is not "overcompensation" because that is made up and makes no sense.
The answer is not "causality" nor "spuriousness." These are strengths of lab experiments. By controlling for possible confounding variables, lab experiments are able to better establish causality.
A. Response rate
B. Overcompensation
C. Causality
D. Generalizability
E. Spuriousness
Answer: D. Generalizability
"The greatest weakness of laboratory experiments lies in their artificiality." -Babbie Ch. 8
Because lab settings are so controlled and unlike the real world, it is cautionary to use its findings to make generalizations about the entire population. Furthermore, Studies 1 and 2's participants were all undergraduate students of a large, Eastern private university. Study 3, on the other hand, surveyed people of differing age, gender, race, educational attainment, region, as well as whether they were California residents. This gave the study much more external validity because the research subjects were much more representative of the population.
The answer is not "response rate" because, if anything, national surveys would have greater issues of response rate than lab experiments. Those who participate in lab experiments are often individuals who choose to participate for one reason or another (e.g. money or extra credit) so therefore would respond during experimentation. Those who receive surveys, however, do not provide consent before receiving surveys so would be less likely to respond.
The answer is not "overcompensation" because that is made up and makes no sense.
The answer is not "causality" nor "spuriousness." These are strengths of lab experiments. By controlling for possible confounding variables, lab experiments are able to better establish causality.
Willer: OVERDOING GENDER
As one man establishes his masculine standing, he necessarily diminishes the standing of other men.
A) This describes hierarchy of needs
B) This describes hierarchical nature of masculinity
C) This describes hegemonic masculinity
D) None of the above
ANSWER: B
A) Incorrect. This is Maslow's theory in psychology.
B) Correct. Although the hegemonic masculinity is relative depending on the region, a society has a set of particular behaviors that are desirable in men. Thus, creates a social pressure compelling men to strive for this standard which in turn threatens other men's standing in their masculinity.
C) Incorrect. This is a standard, a set of behaviors, that is socially constructed for men.
D) Incorrect. See above.
A) This describes hierarchy of needs
B) This describes hierarchical nature of masculinity
C) This describes hegemonic masculinity
D) None of the above
ANSWER: B
A) Incorrect. This is Maslow's theory in psychology.
B) Correct. Although the hegemonic masculinity is relative depending on the region, a society has a set of particular behaviors that are desirable in men. Thus, creates a social pressure compelling men to strive for this standard which in turn threatens other men's standing in their masculinity.
C) Incorrect. This is a standard, a set of behaviors, that is socially constructed for men.
D) Incorrect. See above.
Quiz Question: connecting Willer's article to chapter 8
By pre-sealing the envelopes containing the respondents results to the survey distributed by Willer's research assistants, the possible biases of this experiment were increased/decreased, through the method of:
A. randomization
B. matching
C. double-blind
D. placebo effect
Answers:
DECREASED BIASES through...
A. Randomization: INCORRECT randomization is a way to organize your sample in an experiment, it helps to increase the reliability of your results by getting rid of any trends or commonalities among respondents that wold sway your test results any which way.
B. Matching: INCORRECT matching is also a way of organizing your test subjects to increase the reliability and validity of your test. When a researcher uses matching they pair up their respondents by similarities in important traits such as gender or height (depending on the subject under study) and put one in the control group and one in the experimental group, this way the researchers can see if the dependent variable being tested is actually having an effect on the individual in the experimental group and not the control group.
C. Double-blind- CORRECT double blind tests are a strategy used by researchers to eliminate possible biases that might arise when the researchers know information about the results or conditions of their respondents within an experiment. By pre-sealing the result envelopes the research assistants had no way of knowing which respondents would be assigned as more feminine or more masculine and could therefore not treat them differently in any way that would change the respondents behavior.
D. placebo effect INCORRECT the placebo effect refers to an experiment where the control group is given a fake stimulus, such as a sugar pill, that can not possibly have any impact on them that is not a psychological result of believing that they are being given an active stimulus.
A Test of the Masculine Overcompensation Thesis
(i)What does the masculine overcompensation thesis assert? (ii)What measures were investigated? (iii) What role did testosterone play in masculine overcompensation?
(i)The masculine overcompensation thesis asserts that men react to masculinity threats with extreme demonstrations of masculinity.
(ii)Iraq war support, negative views of homosexuality, and support for then president Bush
(iii)Found no evidence that testosterone mediated the masculine overcompensation effect, consistent support was found for a moderation effect.
(i)The masculine overcompensation thesis asserts that men react to masculinity threats with extreme demonstrations of masculinity.
(ii)Iraq war support, negative views of homosexuality, and support for then president Bush
(iii)Found no evidence that testosterone mediated the masculine overcompensation effect, consistent support was found for a moderation effect.
Key Points/Applications in Willer
It's a pdf.
[UPDATE]: Gdocs' preview function is being rather uncooperative, or I'm just user failing. (Probably the latter.)
If it doesn't open here in Google, try viewing it in Microsoft Docs BETA.
And if THAT doesn't work, here's a question:
WHAT does the step-wise process detailed in this blog post represent?
a) contiguity
b) incompetence
c) statistical regression
d) contingency
[UPDATE]: Gdocs' preview function is being rather uncooperative, or I'm just user failing. (Probably the latter.)
If it doesn't open here in Google, try viewing it in Microsoft Docs BETA.
And if THAT doesn't work, here's a question:
WHAT does the step-wise process detailed in this blog post represent?
a) contiguity
b) incompetence
c) statistical regression
d) contingency
Monday, October 25, 2010
Chapter 9 Quiz Question
Have you ever had asian food?
a. yes
b. no
If so, what type of food?
a. Chinese
b. Korean
c. Vietnamese
d. Thai
What type of question is this?
a. Bias
b. Matrix
c. Contingency
This type of question is a Contingency question because it offers a structure for exploring not only if they eat asian food but in more depth as to what type of asian food they eat.
Babbie Ch.9 Question
What type of question depends on the respondents answer to the first question?
For example, "Have you ever smoked marijuana?"
YES
NO
If yes: About how many times have you smoked marijuana?
Once
2 to 5 times
6 to 10 times
11 to 20 times
More than 20 times
A. Matrix
B. Contingency
Answer: Contingency
In this case, you would not want to ask how many times a person has smoked marijuana as your second question, if in their first question they stated that they had never smoked the substance before. This would be an irrelevant question for them to answer.
Babbie Ch.9 Question
True or False:
Answer: FALSE.
Open-ended questions do question respondents to provide his or her own answer, but also, they require in-depth, qualitative interviewing – unlike close-ended questions. In close ended questions, they rely on survey research because they provide a greater uniformity of respondents and are more easily processed than open-ended questions.
Chapter 9
The number of people who respond to a survey divided by the number of people that were selected to do the survey, as a percent is the __________.
A) Response Rate
B) Completion Rate
C) Return Rate
D) All of the above
E) None
CORRECT ANSWER: D
Because they all mean the same thing: the percent of those individuals who are actually participating over the total surveys sent out to a sample. For instance if we were to do a survey and we sent out the survey to a sample of 200 people but only got back 165 responses out response rate/completion rate/return rate would be 165 / 200 = 82.5%.
Babbie, Ch. 9 Question
According to Babbie, chapter 9, the construction of close-ended questions should be guided by these two structural requirements:
a) The response categories should be clear and unambiguous
b) The response categories should be exhaustive and mutually exclusive
c) The response categories should be exhaustive and unambiguous
d) Survey questions should never be close-ended
Correct answer: b)
a) While true, these two factors are not the structural guidelines Babbie talks about
c) Exhaustiveness is a structural guideline but ambiguity is not mentioned as the second guideline
d) Close-ended questions are widely used and popular because of their uniformity and the ease with which they can be processed
Friday, October 22, 2010
Chap 9 Survey Research
There are multiple types of questions. A few are provided below. Please match the example question with the best type. Some may overlap.
a. contingency question
b. open-ended question
c. closed-ended question
d. double-barreled questions
1. How satisfied are you with the school conditions and school tuition?
Very satisfied, satisfied, not satisfied.
2. What do you feel is the most important issue facing American teenagers today?
3. Do you drink coffee?
Yes, No. If yes, how many cups a day do you drink? One, 2-5, 6-10, 11-20, more then 20
4. Do you know your height?
Yes, No, I dont know
1. d) double-barreled question- A question that asks for a single answer but has multiple parts. This example is asking about the respondents satisfaction towards their school conditions and combining it with another part, their satisfaction with school tuition. These are two parts, combined into one question, forcing the respondent to answer both with one response. (Babbie p 247) This is also an example of a (c) closed ended question, but since it is double barreled, d) would be the best choice, because this is a situation researchers should avoid when creating closed-ended questions.
2. b) open-ended question - A question where the respondent is asked to provide his or her own answers. (Babbie p 246)
3. a) contingency question - A survey question intended for only some respondents, determined by their responses to some other question. (Babbie p 253) This is also (c) a close-ended question, since it provides answers constructed by the researcher for the respondent to choose from.
4. c) close-ended question - Survey quetions in which respondent is asked to select an answer from among a list provided by the researcher. (Babbie 247)
a. contingency question
b. open-ended question
c. closed-ended question
d. double-barreled questions
1. How satisfied are you with the school conditions and school tuition?
Very satisfied, satisfied, not satisfied.
2. What do you feel is the most important issue facing American teenagers today?
3. Do you drink coffee?
Yes, No. If yes, how many cups a day do you drink? One, 2-5, 6-10, 11-20, more then 20
4. Do you know your height?
Yes, No, I dont know
1. d) double-barreled question- A question that asks for a single answer but has multiple parts. This example is asking about the respondents satisfaction towards their school conditions and combining it with another part, their satisfaction with school tuition. These are two parts, combined into one question, forcing the respondent to answer both with one response. (Babbie p 247) This is also an example of a (c) closed ended question, but since it is double barreled, d) would be the best choice, because this is a situation researchers should avoid when creating closed-ended questions.
2. b) open-ended question - A question where the respondent is asked to provide his or her own answers. (Babbie p 246)
3. a) contingency question - A survey question intended for only some respondents, determined by their responses to some other question. (Babbie p 253) This is also (c) a close-ended question, since it provides answers constructed by the researcher for the respondent to choose from.
4. c) close-ended question - Survey quetions in which respondent is asked to select an answer from among a list provided by the researcher. (Babbie 247)
Wednesday, October 20, 2010
Chapter 8 quiz question
Which of the following statements is false?
A. The static group comparison is not a recommended experiment design because it uses pretest methods
B. The static group comparison is not a recommended experiment design because it is seen as less than professional
C. The classical experiment design looks at all 12 forms of internal validity
D. The classical experiment runs the risk of influencing the subjects, or effecting their answers.
E. All of the above statements are true.
Answer: A. In fact the static group method does not use pretest methods.
B. The static group comparison, along with the one group pretest posttest design and the one shot case study are all frequently used in less-than-professional experiments
C. If done correctly the classical experiment design can in fact deal with all twelve point of possible internal validity
D. One of the problems with classical experiment design is that the pretesting runs the risk of influencing the subject, so the statement is true.
Quiz Question ch. 8 Experiments
The degree of resemblance between sample groups in an experiment achieved by probability sampling is largely a function of _______.
A. sample Size
B. randomness
C. matching
D. validity
A. CORRECT - A probability sample of less than 100 is not likely to be representative of the sampling frame we have derived from the population under study. It is important in probability sampling that the groups represent the sampling frame, and therefore it is not good choice to use probability sampling when dealing with a small amount of test subjects.
B. INCORRECT- randomness is not an immediate concern of probability sampling, the main concern is that the samples represent the population under study which can be achieved through a large sample size.
C. INCORRECT- matching is a way to achieve comparability between the experimental group and the control group and is not the main function of probability sampling.
D. INCORRECT - Validity refers to the extent to which we are actually testing what we want to test in our experiment. It is of the utmost important that all studies are valid but it is not what determined whether or not probability sampling will be effective.
Quiz Question Ch. 8
In an experiment, when assigning a subject to a control or experimental group which method would be most effective in studying a large population?
A. Probability Sampling
B. Randomization
C. Matching
D. Web-Based Sampling
A. Nope. In social scientific experiments many subjects are not involved, and probability sampling is not fairly representative if there is less than 100 subjects. This means that probability sampling is seldom used in large populations.
B. THIS IS CORRECT. Randomization works with large populations because each sample (control and experimental) reflects the characteristics of the total population. This means that the two samples will reflect each other.
C. Wrong. Matching is a way to achieve comparability between the control and experimental groups . When separating the groups into sample and control through the process of matching you may not always know what variables will be relevant for the matching process.
D. Wrong, because I made up the term and it just doesn’t make sense.
Tuesday, October 19, 2010
An experiment where neither the subjects nor the experimenter know which is the control group or which is the experimental group is called a:
A) Natural Experiment
B) Randomized Experiment
C) Double Blind Experiment
D) Control Experiment
A) Wrong Answer: A natural experiment can occur in the course of natarul events (like a hurricane) it contains neither a predertemined control or experimental group.
B) Wrong Answer: I made up Randomized Experiment.
C) Correct!! A Double blind experiment is designed to avoid experimenter bias. To ensure that both the experimental and the control group are observed and measured as identically as possible, neither the experimenter nor the subjects know which is the control group or the experimental group.
D) Wrong Answer. I made up up the term: Control Experiment.
A) Natural Experiment
B) Randomized Experiment
C) Double Blind Experiment
D) Control Experiment
A) Wrong Answer: A natural experiment can occur in the course of natarul events (like a hurricane) it contains neither a predertemined control or experimental group.
B) Wrong Answer: I made up Randomized Experiment.
C) Correct!! A Double blind experiment is designed to avoid experimenter bias. To ensure that both the experimental and the control group are observed and measured as identically as possible, neither the experimenter nor the subjects know which is the control group or the experimental group.
D) Wrong Answer. I made up up the term: Control Experiment.
Quiz Question Babbie Ch. 8 Experiments-Subject Selection
Which of the following methods would NOT be chosen by a researcher to select subjects for an experiment if he or she wanted to acheive comparibility in the experiment and control groups?
A. Probability Sampling
B. Double-Blind Selection
C. Matching
D. Randomization
e. All of the above are methods that might be chosen
A. is not correct, because probability sampling is a method of subject selection. it is a general term for samples selected in accord with the probability theory, such as systematic sampling, that typically involves some sort of random selection mechanism.
B. is the CORRECT answer. double-blind selection is not a term used in regards to subject selection. a double-blind experiment, however, is a design implemented once subjects are already chosen to help guard against experimenter bias, where neither the experimenter or the subject knows which subjecs are in the control and experimental groups.
C. is not correct, because matching is a method of subject selection whereby pairs of subjects are matched on the basis of their similarities on one or more variables; one member of the pair is assigned to the control group and the other is assigned to the experimental group.
D. is not correct, because randomization is a method of subject selection. it is the technique of assigning experimental subjects to experimental and control groups randomly.
E. is not correct because double-blind selection is not a method of subject selection.
A. Probability Sampling
B. Double-Blind Selection
C. Matching
D. Randomization
e. All of the above are methods that might be chosen
A. is not correct, because probability sampling is a method of subject selection. it is a general term for samples selected in accord with the probability theory, such as systematic sampling, that typically involves some sort of random selection mechanism.
B. is the CORRECT answer. double-blind selection is not a term used in regards to subject selection. a double-blind experiment, however, is a design implemented once subjects are already chosen to help guard against experimenter bias, where neither the experimenter or the subject knows which subjecs are in the control and experimental groups.
C. is not correct, because matching is a method of subject selection whereby pairs of subjects are matched on the basis of their similarities on one or more variables; one member of the pair is assigned to the control group and the other is assigned to the experimental group.
D. is not correct, because randomization is a method of subject selection. it is the technique of assigning experimental subjects to experimental and control groups randomly.
E. is not correct because double-blind selection is not a method of subject selection.
Monday, October 18, 2010
Question on Chapter 7: The Logic of Sampling – Probability sampling
Which one of the following definitions refers to PPS (a type of probability sampling)?
- A type of sampling in which the units composing a population are assigned numbers. A set of random numbers is then generated, and the units having those numbers are included in the sample.
- A type of sampling in which every kth unit in a list is selected for inclusion in the sample.
- A type of sampling in which each member of population has the same chance of being selected into the sample.
- A type of multistage cluster sample in which clusters are selected, not with equal properties but with probabilities proportionate to their sizes – as measured by the number of units to be subsampled.
- None of the above.
Answer: D
PPS: Probability Proportionate to Size is a type of multistage cluster sample in which clusters are selected, not with equal properties but with probabilities proportionate to their sizes – as measured by the number of units to be subsampled.
EPSEM: A type of multistage cluster sample in which clusters are selected, not with equal properties but with probabilities proportionate to their sizes – as measured by the number of units to be subsampled.
Systematic sampling: A type of sampling in which every kth unit in a list is selected for inclusion in the sample.
Simple random sampling: A type of sampling in which the units composing a population are assigned numbers. A set of random numbers is then generated, and the units having those numbers are included in the sample.
Creating Effective Tables
Which one of the following accurately describes univariate tables most accurately?
a. tables that show the relationship between two variables
b. tables that show give information on one variable its association among other variables
c. tables that present among three variables or sets of related variables, such as joint distribution of three categorical variables
d. tables that show information on each variable alone rather than associations among variables
e. none of the above
a. incorrect-tables that show the relationship between two variables are examples of bivariate tables
b. incorrect-tables that show give information on one variable its association among other variables; univariate tables don't show association among variables
c. incorrect-tables that present among three variables or sets of related variables, such as joint distribution of three categorical variables are three-way tables
d. correct-tables that show information on each variable alone rather than associations among variables are univariate tables. Common types of univariate tables include those that present the distribution of a variable or composition of a sample.
e. incorrect answer-none of the above
a. tables that show the relationship between two variables
b. tables that show give information on one variable its association among other variables
c. tables that present among three variables or sets of related variables, such as joint distribution of three categorical variables
d. tables that show information on each variable alone rather than associations among variables
e. none of the above
a. incorrect-tables that show the relationship between two variables are examples of bivariate tables
b. incorrect-tables that show give information on one variable its association among other variables; univariate tables don't show association among variables
c. incorrect-tables that present among three variables or sets of related variables, such as joint distribution of three categorical variables are three-way tables
d. correct-tables that show information on each variable alone rather than associations among variables are univariate tables. Common types of univariate tables include those that present the distribution of a variable or composition of a sample.
e. incorrect answer-none of the above
Question on Chapter 7: The Logic of Sampling
A ___________ sampling is a nonprobability sampling method often employed in field research whereby each person interviewed may be asked to suggest additional people for interviewing, whereas, ____________ sampling is a type of nonprobability sampling in which units are selected into a sample on the basis of prespecified characteristics, so that the total sample will have the same distribution of characteristics assumed to exist in the population being studied.
- Reliance on variable subjects, Quota
- Purposive, Reliance on variable subjects
- Snowball, Quota
- Quota, Purposive
- Snowball, Purposive
Answer: C
Reliance on available subjects sampling: such as stopping people at a street corner or some other location, is an extremely risky sampling method.
Purposive Sampling: a type of nonprobability sampling in which the units to be observed are selected on the basis of the researcher’s judgment about which ones will be the most useful or representative.
Snowball Sampling: a nonprobability sampling method often employed in field research whereby each person interviewed may be asked to suggest additional people for interviewing
Quota Sampling: is a type of nonprobability sampling in which units are selected into a sample on the basis of prespecified characteristics, so that the total sample will have the same distribution of characteristics assumed to exist in the population being studied.
Question on Creating Effective Tables
Which one of the following definitions fits the Bivariate type of tables?
- Tables showing information on each variable alone rather than associations among variables.
- Tables showing the relationship between two variables.
- Tables presenting information on associations among three variables or sets of related variables.
- Both B and C fall under the Bivariate table type.
Answer: B
Tables showing information on each variable alone rather than associations among variables are called Univariate Tables.
Tables showing the relationship between two variables are Bivariate.
Tables presenting information on associations among three variables or sets of related variables are Three-Way Tables.
Because B and C have their own definitions they don’t fit into a single category.
Sunday, October 17, 2010
Babbie Chp. 7 Nonprobability Sampling
Convenience or haphazard sampling is another name for what type of nonprobability sampling technique?
A) reliance on available subjects
B) purposive or judgmental sampling
C) snowball sampling
D) quota sampling
E) none of the above
A) Correct! Relying on available subjects is sometimes called convenience or haphazard sampling. An example of reliance on available subjects is stopping people at a street corner.
B) Wrong! Purposive or judgmental sampling is a type of nonprobability sampling in which the units to be observed are selected on the basis of the researcher's judgment about which ones will be the most useful or representative. This would be selecting a sample on the basis of knowledge of a population.
C) Wrong! Snowball sampling is a nonprobability sampling method often employed in field research whereby each person interviewed may be asked to suggest additional people for interviewing. It is also known as a form of accidental sampling. This is used when members of a special population are difficult to locate, like homeless.
D) Wrong! Quota sampling is a type of nonprobability sampling in which units are selected into a sample on the basis of prespecified characteristics, so that the total sample will have the same distribution of characteristics assumed to exist in the population being studied.
E) Wrong! Reliance on available subjects is the correct answer!
A) reliance on available subjects
B) purposive or judgmental sampling
C) snowball sampling
D) quota sampling
E) none of the above
A) Correct! Relying on available subjects is sometimes called convenience or haphazard sampling. An example of reliance on available subjects is stopping people at a street corner.
B) Wrong! Purposive or judgmental sampling is a type of nonprobability sampling in which the units to be observed are selected on the basis of the researcher's judgment about which ones will be the most useful or representative. This would be selecting a sample on the basis of knowledge of a population.
C) Wrong! Snowball sampling is a nonprobability sampling method often employed in field research whereby each person interviewed may be asked to suggest additional people for interviewing. It is also known as a form of accidental sampling. This is used when members of a special population are difficult to locate, like homeless.
D) Wrong! Quota sampling is a type of nonprobability sampling in which units are selected into a sample on the basis of prespecified characteristics, so that the total sample will have the same distribution of characteristics assumed to exist in the population being studied.
E) Wrong! Reliance on available subjects is the correct answer!
Saturday, October 16, 2010
Chapter 7 Quiz Question - Measuring Accuracy of Samples
Nancy wants to determine how accurate her sample statistic is from the parameter of the population she is studying. Which of the following should she calculate to obtain this information?
a) standard deviation
b) sampling error
c) confidence level
d) confidence interval
Answers:
a) incorrect - Standard deviation is "a measure of dispersion around the mean" (414). It is not directly related to the accuracy of of a sample statistic from the parameter of a population. However, one should note that "standard error is the standard deviation of the sampling distribution" (Babbie 196).
b) correct - Sampling error is "the degree of error to be expected in probability sampling", which is what Nancy wants to know (196).
c) incorrect - Confidence level, a component of sampling error, is is "the estimated probability that a population parameter lies within a given confidence interval" (197). While Nancy would want to know this information, it would not tell her everything she needs to know about the accuracy of her sample statistic.
d) incorrect - Confidence interval, a component of sampling error, is "the range of values within which a population value is estimated to lie" (197). While Nancy would want to know this information, it would not tell her everything she needs to know about the accuracy of her sample statistic.
a) standard deviation
b) sampling error
c) confidence level
d) confidence interval
Answers:
a) incorrect - Standard deviation is "a measure of dispersion around the mean" (414). It is not directly related to the accuracy of of a sample statistic from the parameter of a population. However, one should note that "standard error is the standard deviation of the sampling distribution" (Babbie 196).
b) correct - Sampling error is "the degree of error to be expected in probability sampling", which is what Nancy wants to know (196).
c) incorrect - Confidence level, a component of sampling error, is is "the estimated probability that a population parameter lies within a given confidence interval" (197). While Nancy would want to know this information, it would not tell her everything she needs to know about the accuracy of her sample statistic.
d) incorrect - Confidence interval, a component of sampling error, is "the range of values within which a population value is estimated to lie" (197). While Nancy would want to know this information, it would not tell her everything she needs to know about the accuracy of her sample statistic.
Friday, October 15, 2010
Chapter 16 Question
A difference between descriptive statistics and inferential statistics is:
A) both summarize a set of sample observations
B) both move beyond the description of specific observations to make inferences about the larger population sample
C) the former summarizes the sample observations while the latter goes beyond and infers broadly from which the sample observations were drawn.
Answer: c)
A) both summarize a set of sample observations
B) both move beyond the description of specific observations to make inferences about the larger population sample
C) the former summarizes the sample observations while the latter goes beyond and infers broadly from which the sample observations were drawn.
Answer: c)
Thursday, October 14, 2010
Ch. 16 Quiz question: Measures of association
The following levels of measurement uses which measure of association respectively: nominal, ordinal, interval and ratio?
A. Gamma, Sigma, Lambda
B. Gamma, Lambda, Pearson's r
C. Lambda, Gamma, Perason's r
D. Lambda, Gamma, Alpha
E. Alpha, Gamma, Sigma
Answer: C
For nominal variables, the lambda measure of association is used (however, it is not limited to lambda - lambda is only one of several measure of association that can be used to analyze two nominal variables) Lambda is based on your ability to guess the values of one variable through knowing the values of another variable. Values of lambda vary from 0 to 1.
For ordinal variables, gamma is one appropriate measure of association. Like lambda, gamma is based on our ability to guess values on one variable by knowing values on another. However, gamma is based on guessing the ordinal arrangement of values. Values of gamma vary from -1 to +1, representing the direction as well as the magnitude of the association (positive or negative correlation).
For interval and ratio variables, Pearson’s product-moment correlation (r) is one appropriate measure of association. Like both gamma and lambda, r is based on guessing the value of one variable by knowing another. However, more specifically, r reflects how closely you can guess the value of one variable through your knowledge of the value of another.
Sigma and alpha are made up and irrelevant in this case. They were not in the reading.
A. Gamma, Sigma, Lambda
B. Gamma, Lambda, Pearson's r
C. Lambda, Gamma, Perason's r
D. Lambda, Gamma, Alpha
E. Alpha, Gamma, Sigma
Answer: C
For nominal variables, the lambda measure of association is used (however, it is not limited to lambda - lambda is only one of several measure of association that can be used to analyze two nominal variables) Lambda is based on your ability to guess the values of one variable through knowing the values of another variable. Values of lambda vary from 0 to 1.
For ordinal variables, gamma is one appropriate measure of association. Like lambda, gamma is based on our ability to guess values on one variable by knowing values on another. However, gamma is based on guessing the ordinal arrangement of values. Values of gamma vary from -1 to +1, representing the direction as well as the magnitude of the association (positive or negative correlation).
For interval and ratio variables, Pearson’s product-moment correlation (r) is one appropriate measure of association. Like both gamma and lambda, r is based on guessing the value of one variable by knowing another. However, more specifically, r reflects how closely you can guess the value of one variable through your knowledge of the value of another.
Sigma and alpha are made up and irrelevant in this case. They were not in the reading.
Wednesday, October 13, 2010
Study tool: use of units and labeling in effective tables (Effective Tables, Jane E. Miller)
When making an effective table it is important to identify the units being measured at several steps. Rather than just identifying the unit of analysis like we have done previously when analyzing research studies, when making tables it is necessary to identify and label several different units and their specific measurement within one graph.
in a table you must identify these for all variables:
-units of measurement
-level of aggregation
-system of measurement
Than appropriate place to identify these variables/units is
1. in the title
2. row labels
3. column headings
Examples:
Title: number and type of fast food restaurants by every 100 square miles, California, 2010
Row labels: fast food restaurants (concept), Burger King, McDonald's, Taco Bell (attributes)
Column headings: area 1, area 2, area 3, area 4, total.
Here we have identified that the unit of analysis is institutions (fast food restaurants), the measure is by every 100 square miles, and that the concept of fast food restaurants that we are conceptualizing for our purposes includes, Burger King, Taco Bell, and McDonald's.
Memo - Babbie Chapter 16: Statistical Analysis
Chapter 16: Statistical Analyses
What is Descriptive Statistics (DS), and what is its objective?
- Descriptive statistics is a statistical computation describing ether the characteristics of a sample or the relationship among variables in a sample.
- Objective: DS summarizes a set of sample observations or the association between two variables into measures of proportionate reduction of error (PRE)
PRE – A logical model for assessing the strength of a relationship by asking how much knowing values on one variable would reduce our errors in guessing values in the other.
There are 3 models of PRE:
Nominal Variables – if two variables consist of nominal data (ex. Gender, religious affiliation, race) you can substitute each variable with a lambda (λ) to measure the exact value in proportion to the overall distribution.
Example: Of the group of 500 students, who attended the conference, 375 students are supporters of the Republican Party and 125 students were supporters of the Democratic Party. If you were to have guessed how many students are affiliated with the Republican Party or the Democratic Party, the exact value in proportion to the overall distribution of Democratic Students will be .25. λ = .25.
Ordinal Variables – Ordinal data (social class, religiosity, and alienation) measure a relationship between two variables, and uses the symbol gamma (γ) to guess values on one variable by knowing values on another.
Example from the book: Check for correlation. Are the variables showing a negative or positive relationship?
Let’s say you suspect that religiosity is positively related to political conservatism, and if Person A is more religious than Person B, you guess that A is also more conservative than B. In a negatively related correlation, Person A is more religious than Person B, but A is less conservative than B.
Interval / Ratio Variables – one measurement for interval/ratio variables is Pearson’s product-moment correlation (r), which is based on guessing the value of one variable by knowing another, or how closely you can guess the value of one variable through your knowledge of the value of another. For continuous interval or ratio variables, it is unlikely that you could predict the precise value of the variable.
Monday, October 11, 2010
Study Tool: Measurement Quality Criteria (Babbie Ch. 5, Lecture 9)
Describe how the four measurement quality criteria (accuracy, reliability, validity, precision) relate to scenarios A through D. Consider which can or cannot be explained in each scenario. All four concepts will be used, though some will apply to more scenarios than others.
Scenarios:
A) A scale that is always 5 pounds off.
B) Using a scale to measure temperature.
C) A scale that gives a different weight every time.
D) A scale that tells the weight of something to the nearest hundredth of a pound.
Definitions to keep in mind:
- precision: how concise a measure is.
- validity: the degree to which a measure reflects the underlying construct
- reliable: whether or not the same measure can be produced if the test is carried out multiple times.
- accuracy: whether or not a measure reflects reality.
Answers May Vary - Example Answers
A) Since the scale is 5 pounds off, it is not accurate. However, since it is always 5 pounds off it is reliable.
B) A scale will not give a valid measure of temperature, since scales measure weight.
C) A scale that gives a different weight ever time is not reliable, nor is it accurate.
D) A scale that gives the weight of an object to the nearest hundredth of a pound is very precise. Just because the scale is precise, we cannot assume it is also accurate, valid, or reliable.
Scenarios:
A) A scale that is always 5 pounds off.
B) Using a scale to measure temperature.
C) A scale that gives a different weight every time.
D) A scale that tells the weight of something to the nearest hundredth of a pound.
Definitions to keep in mind:
- precision: how concise a measure is.
- validity: the degree to which a measure reflects the underlying construct
- reliable: whether or not the same measure can be produced if the test is carried out multiple times.
- accuracy: whether or not a measure reflects reality.
Answers May Vary - Example Answers
A) Since the scale is 5 pounds off, it is not accurate. However, since it is always 5 pounds off it is reliable.
B) A scale will not give a valid measure of temperature, since scales measure weight.
C) A scale that gives a different weight ever time is not reliable, nor is it accurate.
D) A scale that gives the weight of an object to the nearest hundredth of a pound is very precise. Just because the scale is precise, we cannot assume it is also accurate, valid, or reliable.
Babbie, Ch. 10 Midterm Question
Which of the following is NOT TRUE about indexes?
a) indicators combined in an index are given equal weight
b) Subjects are scored on the basis of their overall pattern of responses
c) Subjects are scored on the basis of their total number of positive (or negative) responses
d) The Gender Empowerment Measure is an example of an index
ANSWER:
b)
In an scale, subjects are scored on the basis of their overall pattern of responses. Indexes are scored on the basis of their total number of positive (or negative) responses
Sunday, October 10, 2010
Babbie Chp. 6: Typologies Midterm Quiz Question
Which of the following statements is true about typologies?
A) Typologies should always be avoided in social research
B) It is extremely difficult to analyze a typology as an independent variable
C) It is extremely difficult to analyze a typology as a dependent variable
D) A typology is an ordinal composite measure
E) none of the above statements are true
A) False! Don't think that typologies should always be avoided in social research. Often they provide the most appropriate device for understanding the data.
B) False! Whenever the typology is used as the independent variable, there will probably be no problem.
C) Correct! It is true that it is hard to analyze a typology as a dependent variable.
D) False! A typology is a nominal composite measure often used in social research.
E) False! One of the answers, C, is correct.
A) Typologies should always be avoided in social research
B) It is extremely difficult to analyze a typology as an independent variable
C) It is extremely difficult to analyze a typology as a dependent variable
D) A typology is an ordinal composite measure
E) none of the above statements are true
A) False! Don't think that typologies should always be avoided in social research. Often they provide the most appropriate device for understanding the data.
B) False! Whenever the typology is used as the independent variable, there will probably be no problem.
C) Correct! It is true that it is hard to analyze a typology as a dependent variable.
D) False! A typology is a nominal composite measure often used in social research.
E) False! One of the answers, C, is correct.
Identifying Moderator and Mediator Variables
1) The negative impact of heavy drinking on the grade point averages of college students is accentuated by the daily use of cigarettes.
Independent variable?
Dependent variable?
Name the moderator/mediator, and explain your answer.
ANSWER:
Independent variable: level of drinking
Dependent variable: GPAs of college students
Moderator, daily cigarette smoking. The effect of smoking cigarettes accentuates the effects of heavy drinking, but does not by itself cause lower GPAs.
CH. 16 Babbie, Inferential Statistics Memo
- inferential statistics are different than descriptive statistics in that they don't merely describe a sample with regression analysis, but they use a sample to make estimates about a larger population. in order for various inferential statistical calculations to be relevant, they must follow a few conditions...
1) the sample from which statistical measures are taken must be drawn from the population about which inferences are being made.
2) the statistics assume sampling occurred with replacement, assume simple random sampling occurred in the sample, and assume the sample had a 100% completion rate. if these assumptions don't hold true, it is not usually a serious problem, as long as the sample remains representative of the entire population.
3) the statistics address sampling error only, and not nonsampling error (the imperfections of data quality that are due to things such as misunderstandings of questions by respondents or coding errors). the calculated level of statistical significance depends on how likely the relationships observed in the sample are due to sampling error alone. however, since nonsampling error is often larger than sampling error, generalizing sample findings to a population must be done with caution.
- tests of statistical significance are statistical computations that indicate the likelihood that the relationship observed between variables in a sample can be attributed to sampling error (unrepresentativeness) only. statistical significance is always expressed in probabilities. the probability of the measured associations being due only to sampling error is called the level of significance. some researchers specify in advance the level of significance they will regard as sufficient for the purposes of their test, and if the measured association is statistically significant at that level, they'll regard it as representing a genuine association between the two variables.
- statistical significance is not to be confused with SUBSTANTIVE significance, which means that an observed association is strong, important, meaningful, etc.
1) the sample from which statistical measures are taken must be drawn from the population about which inferences are being made.
2) the statistics assume sampling occurred with replacement, assume simple random sampling occurred in the sample, and assume the sample had a 100% completion rate. if these assumptions don't hold true, it is not usually a serious problem, as long as the sample remains representative of the entire population.
3) the statistics address sampling error only, and not nonsampling error (the imperfections of data quality that are due to things such as misunderstandings of questions by respondents or coding errors). the calculated level of statistical significance depends on how likely the relationships observed in the sample are due to sampling error alone. however, since nonsampling error is often larger than sampling error, generalizing sample findings to a population must be done with caution.
- tests of statistical significance are statistical computations that indicate the likelihood that the relationship observed between variables in a sample can be attributed to sampling error (unrepresentativeness) only. statistical significance is always expressed in probabilities. the probability of the measured associations being due only to sampling error is called the level of significance. some researchers specify in advance the level of significance they will regard as sufficient for the purposes of their test, and if the measured association is statistically significant at that level, they'll regard it as representing a genuine association between the two variables.
- statistical significance is not to be confused with SUBSTANTIVE significance, which means that an observed association is strong, important, meaningful, etc.
Wednesday, October 6, 2010
Oct 6th and 7th -- letter from Tom Pessah
A letter from my colleague Tom Pessah
Tom
~~~~~~~~~~~~~~~~~
Dear Students,
Sociology as a discipline, and this department in particular, tends to engage in social struggles (see for example http://www.walmartmovie.com/
The campus does, of course, have a rich history of protests on a whole range of issues - http://caldisorientation.
as you can see in this movie - http://www.fandango.com/
available from the Media Resources Center in Moffit).
The protests this week are a part of a national day of action that will include mass protests, rallies, walkouts, strikes, and sit-ins, against the attacks on public education and public services. YOUR tuition increased last year by 32%. There are plans to eliminate 200 campus faculty and staff positions, Chicano Studies and Asian American Studies as majors may disappear, and there’s been a 12% drop in Latino/a admissions.
At their last meeting, the UC Regents voted unanimously to cut pensions for the UC’s lowest paid workers and to increase the pensions of the UC’s 250 highest paid employees. This news comes only a few short weeks after the New York Times and other major news agencies reported that, before moving to his new mansion in Lafayette, UC President Mark Yudof racked up $70,000 worth of damages to his previous UC mansion.
The UC master plan of 1960 affirms the right of every Californian to affordable education of the highest quality. Higher education in California used to be almost free.
The is nothing inevitable about these fee hikes, it's a matter of priorities. Three decades ago, 10.1 percent of the state's general fund was allocated to higher education, 3.4 percent to prisons. Today, prisons receive almost 11 percent of the budget, whereas higher education has dropped to 7.5 percent.http://www.today.ucla.edu/
As a direct result of last years' protests, the Governor has already proposed to restore some the UC's funding, although it isn't clear if this will get approved
Among the actions planned this week are the Wed 10/6 TEACH-IN SPEAK-OUT at 5:30pm (7th floor of Eshelmann Hall on Banrcroft) on The Looting of our University and other topics,
and the Thu 10/7 RALLY at Sproul Plaza at 12 PM. There will be events all day on Thursday.
Tom
Tuesday, October 5, 2010
Chapter 16 Quiz Question
Q: Which of the following should NOT be assumed when performing regression analyses?
a) continuous interval data
b) extrapolation of data
c) absence of nonsampling errors
d) simple random sampling
A: b) extrapolation of data
Evidence: In the text, Babbie writes "The use of regression analysis for statistical inferences is based on... simple random sampling, the absence of nonsampling errors, and continuous interval data," which covers the answers a, c, and d. Babbie, however, warns us about extrapolation, suggesting that some regression equations may not be representative when expanded to population sizes outside their range of observation (see the first paragraph of the second column on pg. 459).
Monday, October 4, 2010
Multivariate Compared with Bivariate and Univariate Analysis
Multivariate analysis is a method of analysing more than two variables. For instance, the independent variables of gender and age can be used to examine their effects on a dependent variable such as yearly income. Not only could you compare the differences of salaries between genders, you can also see how age plays a role in earning capacity within the genders and between them.
Bivariate analysis focuses on only two variables such as gender(IV) and salary(DV) and their relation.
Univariate analysis is employed to examine the dispersion of a specific variable such as gender over an entire population.
Bivariate analysis focuses on only two variables such as gender(IV) and salary(DV) and their relation.
Univariate analysis is employed to examine the dispersion of a specific variable such as gender over an entire population.
Q/A - Chapter 16, type statistical analysis
What type of statistical analysis is this?
The effects of one of more variables are held constant in a statistical analysis. The results may the be compared to the uncontrolled relationship between the two variables to clarify further overall relationships.
EX. We might test the effects of prejudice on education with age held constant.
a. Partial regression analysis
b. Multiple regression analysis
c. Inferential statistics
d. Statistical significance
a. correct. Partial regression analysis. In Babbie " a form of regression analysis in which the effects of one or more variables are held constant, similar to the logic of elaboration model" (458.)
b. incorrect. Multiple regression analysis. In Babbie " A form of statistical analysis that seeks the equation representing the impact of two or more independent variables on a single dependent variable" (458.)
c. incorrect. Inferential statistics. Has nothing to do with the question. It is when "the statistical measures used for making inferences from findings based on sample observations to a larger population" (460.)
d. incorrect. Statistical significance. Has nothing to do with question either. it is a "general term referring to the likelihood that the relationships observed in a sample could be attributed to sample error alone" (461.)
The effects of one of more variables are held constant in a statistical analysis. The results may the be compared to the uncontrolled relationship between the two variables to clarify further overall relationships.
EX. We might test the effects of prejudice on education with age held constant.
a. Partial regression analysis
b. Multiple regression analysis
c. Inferential statistics
d. Statistical significance
a. correct. Partial regression analysis. In Babbie " a form of regression analysis in which the effects of one or more variables are held constant, similar to the logic of elaboration model" (458.)
b. incorrect. Multiple regression analysis. In Babbie " A form of statistical analysis that seeks the equation representing the impact of two or more independent variables on a single dependent variable" (458.)
c. incorrect. Inferential statistics. Has nothing to do with the question. It is when "the statistical measures used for making inferences from findings based on sample observations to a larger population" (460.)
d. incorrect. Statistical significance. Has nothing to do with question either. it is a "general term referring to the likelihood that the relationships observed in a sample could be attributed to sample error alone" (461.)