MSU COM 300 EXAM 2 QUESTIONS
WITH 100% CORRECT ANSWERS
What are the benefits to biased samples? - Answer- -cost
-hard to collect populations
-convenience
-but no theory that indicates that these samples are unbiased
- What are costs associated with biased samples? - Answer- we can only generalize the
population if we have an unbiased sample of that population
-not always obvious what could bias a sample
-best way to avoid a bias, use a systematic method (simple random sample)
Simple Random Sampling - Answer- Get a list of everyone in the population-Generate
the appropriate random numbers.-Select individuals corresponding with your random #'s
Systematic Sampling - Answer- select sample size by determining the sample needed
from the population- list everyone in the population-calculate a skip interval - population
size/sample size -pick a random starting point and skip the interval number-make sure
every person has an equal chance of being selected
ex. particular class with a particular type of major
Stratified Random Sampling - Answer- choose numbers from subgroups that're equal to
their numbers in the population (Population divided into homogeneous sub groups )
Example of Stratified Random Sampling - Answer- (Ex: males and females, use random
selection to ensure the random selection instead of getting males and females)
• Decide the number of people selected from each sub group
• Random sampling
Cluster Sampling - Answer- ♣ Specify groups of individuals you want to sample
♣ Randomly select some of the group
♣ Survey ALL the people in selected group
Example of Cluster Sampling - Answer- Suppose we are interested in surveying people
in Michigan
o Randomly select towns in MI
o Randomly select streets in a selected town
, o Investigate all people who live on the selected street in the selected town
When to use cluster sampling: - Answer- used if no list is available or other methods are
too costly
-specify groups of individuals you wish to sample (dorm, city block)
- randomly select groups
-survey everyone in the selected groups
-Sample everyone on a street/ in a class/ In a building, etc.
Types of Survey Questions - Answer- Semantic differential, likert-type, thermometer,
mC formats, open-ended questions, close-ended questions
Semantic Differential Survey - Answer- has opposite adjectives on the scale and it
chooses the point on the scale that describes something
-Ex: survey items (not likable (1)... likeable (7))
Likert-type Survey - Answer- respondents specify their level of agreement to a
statement
-Ex: "I meet the person often..." (disagree (1) ... Agree (7))
Thermometer Survey - Answer- typically 0 to 100 scale
-Ex: used in marketing/advertising
Multiple Choice Format Surveys - Answer- preconceived questions
Open-ended questions - Answer- given area to write your response
Close-ended questions - Answer- Given a choice of options between answers
Difference between Open-ended and Close-ended questions: - Answer- Open-ended
questions: it's too hard to code
-it's demanding for respondents
-it's hard to compare (long vs. short responses because all responses are different for
each person)
Close-ended questions:
-it can overlook categories
-it's too hard to create relevant responses if unfamiliar with the topic
-its' options limit participants responses
Problems with surveys - Answer- Social desirability, leading phrases, confusing
questions, non-responses
Social Desirability - Answer- Surveys may be inappropriate when dealing with socially
desirable/undesirable behaviors & attitudes because theres a tendency in respondents
to give wrong answers
WITH 100% CORRECT ANSWERS
What are the benefits to biased samples? - Answer- -cost
-hard to collect populations
-convenience
-but no theory that indicates that these samples are unbiased
- What are costs associated with biased samples? - Answer- we can only generalize the
population if we have an unbiased sample of that population
-not always obvious what could bias a sample
-best way to avoid a bias, use a systematic method (simple random sample)
Simple Random Sampling - Answer- Get a list of everyone in the population-Generate
the appropriate random numbers.-Select individuals corresponding with your random #'s
Systematic Sampling - Answer- select sample size by determining the sample needed
from the population- list everyone in the population-calculate a skip interval - population
size/sample size -pick a random starting point and skip the interval number-make sure
every person has an equal chance of being selected
ex. particular class with a particular type of major
Stratified Random Sampling - Answer- choose numbers from subgroups that're equal to
their numbers in the population (Population divided into homogeneous sub groups )
Example of Stratified Random Sampling - Answer- (Ex: males and females, use random
selection to ensure the random selection instead of getting males and females)
• Decide the number of people selected from each sub group
• Random sampling
Cluster Sampling - Answer- ♣ Specify groups of individuals you want to sample
♣ Randomly select some of the group
♣ Survey ALL the people in selected group
Example of Cluster Sampling - Answer- Suppose we are interested in surveying people
in Michigan
o Randomly select towns in MI
o Randomly select streets in a selected town
, o Investigate all people who live on the selected street in the selected town
When to use cluster sampling: - Answer- used if no list is available or other methods are
too costly
-specify groups of individuals you wish to sample (dorm, city block)
- randomly select groups
-survey everyone in the selected groups
-Sample everyone on a street/ in a class/ In a building, etc.
Types of Survey Questions - Answer- Semantic differential, likert-type, thermometer,
mC formats, open-ended questions, close-ended questions
Semantic Differential Survey - Answer- has opposite adjectives on the scale and it
chooses the point on the scale that describes something
-Ex: survey items (not likable (1)... likeable (7))
Likert-type Survey - Answer- respondents specify their level of agreement to a
statement
-Ex: "I meet the person often..." (disagree (1) ... Agree (7))
Thermometer Survey - Answer- typically 0 to 100 scale
-Ex: used in marketing/advertising
Multiple Choice Format Surveys - Answer- preconceived questions
Open-ended questions - Answer- given area to write your response
Close-ended questions - Answer- Given a choice of options between answers
Difference between Open-ended and Close-ended questions: - Answer- Open-ended
questions: it's too hard to code
-it's demanding for respondents
-it's hard to compare (long vs. short responses because all responses are different for
each person)
Close-ended questions:
-it can overlook categories
-it's too hard to create relevant responses if unfamiliar with the topic
-its' options limit participants responses
Problems with surveys - Answer- Social desirability, leading phrases, confusing
questions, non-responses
Social Desirability - Answer- Surveys may be inappropriate when dealing with socially
desirable/undesirable behaviors & attitudes because theres a tendency in respondents
to give wrong answers
