COM3000 Exam 2 Study Guide
Define population - Answer all cases (text/people)
Define sample - Answer a smaller subgroup (of that population)
Define statistic - Answer fact about the sample ex.) their average age of 42
Define parameter - Answer (numerical) fact about the population (whats average, whats
typical about those groups of people)
What are descriptive statistics - Answer Conduct simple descriptions of the
characteristics of a sample (sex, typical age, level in college, number of Facebook
friends- take people/messages and describe them)
What are inferential statistics - Answer Make statements about the characteristics of the
population
What is random sampling - Answer -Equal chance of selection (being chosen for study in
your group)
What is non-random sampling - Answer the samples are gathered in a process that does
not give all the individuals in the population equal chances of being selected.
What are the sub-types of random sampling - Answer 1. Simple Random
-Select at random
-number table, digit dialing, or generator (websites and tools to help you)-
http://www.randomizer.org/form.htm
2. Systematic
-Select every __X__th person (ex. every 10th person) ex.) sometimes can be determined
by the features of your list. if you are doing a study on the population of kalamazoo, and
you have a lot of "s" last names, instead of simple random, it would be best to use
systematic because you would get the correct portion of "s" last names, etc
3. Stratified
-Select at random for known proportions (make sure you have demographic match.
instead of using the whole class randomly, separate men + women and sample the same
amount from each group)
4. Cluster
-Randomly select clustered, then individuals within
What are the sub-types of non-random sampling - Answer Convenience
, -Readily available (showing up at shopping malls, big product fairs, getting people to
take a survey// university researchers = college students)
2. Quota
-Readily available into known population proportions (attitude research about WMU,
important spread of freshman > seniors from campus)
3. Purposive
-Some seem more fit for the research (gonna need to target, ex.) older adults political
attitudes 75+, go to a retirement community)
4. Snowball
-Referrals from initials informants (good in dealing with rare characteristics, ex.) find
couples who had happy divorce and could co-parents well. they probably know other
couples in the same boat. bird of a feather flock together, we hang around others with
same characteristics. you find a few, they find a few)
5. Consecutive (census)
-All accessible participants are chosen
Why should you use random sampling - Answer 1. Less bias
2. More generalizable (closer to population averages)
3. More statical options (for analyzing)
Why should you use non-random sampling - Answer When to use Non-Random Sampling
-To demonstrate a trait (actually) exists (introductory sections of research- this is
helpful)
-To do a qualitative research (qualitative- more verbal, quantitative- numeric) ex. Case
Study, interviews,
-When randomization is not possible
-No list of members
-Very large population
-When generalizing is unimportant (sometimes just want to describe characteristics of
little slice)
-When resources are limited
-As a first step
What is sampling error? - Answer -Always exists
, -The difference between the population mean and the sample mean (how big is the gap?
report this to an audience)
-u and X (meuw and X bar) u- perimeter- fact about the population
X bar- face about the _____
What are steps you can take to decrease sampling error? - Answer Sample:
1. Large
-Central Limit Theorem: the larger the sample size, the more it approximates the
normal distribution (of score)
2. Representative
Calculate Margin of Error
-Confidence Interval
-Range (+ / -) "this is true, + or - 4 points, etc"
Can you ever eliminate sampling error completely? Why or why not - Answer
How do you calculate and interpret the amount of sampling errors? - Answer
What is the basic idea of confidence intervals? What do the ranges mean? - Answer X =
mean (average)
s= Standard deviation
n= sample size
-95%: closer margin, but less certainty
-99%: wider margin, but more certainty
-The larger the sample size, the closer the margin (more ppl you have, the tighter your
margin will be)
Which step of the research process is the time to consider measurement issues -
Answer Operational Definitions
What is an operational definition - Answer The process of determining the observable
characteristics associated with a concept or variable (determining how to
_______________)
In order of least to most informative, what are the four levels of measurement - Answer
ways of assigning numbers to determine the relative amount of something
1. Nominal
Define population - Answer all cases (text/people)
Define sample - Answer a smaller subgroup (of that population)
Define statistic - Answer fact about the sample ex.) their average age of 42
Define parameter - Answer (numerical) fact about the population (whats average, whats
typical about those groups of people)
What are descriptive statistics - Answer Conduct simple descriptions of the
characteristics of a sample (sex, typical age, level in college, number of Facebook
friends- take people/messages and describe them)
What are inferential statistics - Answer Make statements about the characteristics of the
population
What is random sampling - Answer -Equal chance of selection (being chosen for study in
your group)
What is non-random sampling - Answer the samples are gathered in a process that does
not give all the individuals in the population equal chances of being selected.
What are the sub-types of random sampling - Answer 1. Simple Random
-Select at random
-number table, digit dialing, or generator (websites and tools to help you)-
http://www.randomizer.org/form.htm
2. Systematic
-Select every __X__th person (ex. every 10th person) ex.) sometimes can be determined
by the features of your list. if you are doing a study on the population of kalamazoo, and
you have a lot of "s" last names, instead of simple random, it would be best to use
systematic because you would get the correct portion of "s" last names, etc
3. Stratified
-Select at random for known proportions (make sure you have demographic match.
instead of using the whole class randomly, separate men + women and sample the same
amount from each group)
4. Cluster
-Randomly select clustered, then individuals within
What are the sub-types of non-random sampling - Answer Convenience
, -Readily available (showing up at shopping malls, big product fairs, getting people to
take a survey// university researchers = college students)
2. Quota
-Readily available into known population proportions (attitude research about WMU,
important spread of freshman > seniors from campus)
3. Purposive
-Some seem more fit for the research (gonna need to target, ex.) older adults political
attitudes 75+, go to a retirement community)
4. Snowball
-Referrals from initials informants (good in dealing with rare characteristics, ex.) find
couples who had happy divorce and could co-parents well. they probably know other
couples in the same boat. bird of a feather flock together, we hang around others with
same characteristics. you find a few, they find a few)
5. Consecutive (census)
-All accessible participants are chosen
Why should you use random sampling - Answer 1. Less bias
2. More generalizable (closer to population averages)
3. More statical options (for analyzing)
Why should you use non-random sampling - Answer When to use Non-Random Sampling
-To demonstrate a trait (actually) exists (introductory sections of research- this is
helpful)
-To do a qualitative research (qualitative- more verbal, quantitative- numeric) ex. Case
Study, interviews,
-When randomization is not possible
-No list of members
-Very large population
-When generalizing is unimportant (sometimes just want to describe characteristics of
little slice)
-When resources are limited
-As a first step
What is sampling error? - Answer -Always exists
, -The difference between the population mean and the sample mean (how big is the gap?
report this to an audience)
-u and X (meuw and X bar) u- perimeter- fact about the population
X bar- face about the _____
What are steps you can take to decrease sampling error? - Answer Sample:
1. Large
-Central Limit Theorem: the larger the sample size, the more it approximates the
normal distribution (of score)
2. Representative
Calculate Margin of Error
-Confidence Interval
-Range (+ / -) "this is true, + or - 4 points, etc"
Can you ever eliminate sampling error completely? Why or why not - Answer
How do you calculate and interpret the amount of sampling errors? - Answer
What is the basic idea of confidence intervals? What do the ranges mean? - Answer X =
mean (average)
s= Standard deviation
n= sample size
-95%: closer margin, but less certainty
-99%: wider margin, but more certainty
-The larger the sample size, the closer the margin (more ppl you have, the tighter your
margin will be)
Which step of the research process is the time to consider measurement issues -
Answer Operational Definitions
What is an operational definition - Answer The process of determining the observable
characteristics associated with a concept or variable (determining how to
_______________)
In order of least to most informative, what are the four levels of measurement - Answer
ways of assigning numbers to determine the relative amount of something
1. Nominal
