Home
Subjects
Create
Search
Log in
Sign up
Upgrade to remove ads
Only $2.99/month
Exam 2
STUDY
Flashcards
Learn
Write
Spell
Test
PLAY
Match
Gravity
Terms in this set (31)
Nominal Scale
Objects are assigned to mutually exclusive, labeled categories
No necessary relationships among categories-No ordering or spacing are implied
Only possible arithmetic operation is a count of each category
If only two options: binary-Are you a resident of CT? yes / no
Ex. during which season of the year were you born?
Ordinal/Rank Scale
Ranks objects or arranges them in order by some common variable
Does not provide information on how much difference there is between objects
Arithmetic operations are limited to statistics such as median or mode
Preference may correlate with recall
-Ex. Rank your preferences for the following attributes in making a car purchase decision-Price, safety, design, fuel economy
-This assumes that the differences are evenly spaced, 1 point between each →Ex. rank the following according to your preference - tide, surf, cheer, wisk, bold
Constant-sum scale (ordinal)
respondents allocate a fixed number of rating points among serial objects to reflect relative preference
Break preference structure by asking to allocate points out of 100-Nike: 30 / Adidas: 30 / Doc Martin: 40 / Spurga: 0 = 100
Improvement on rank or ordinal scale
Downside: information overload, cognitively taxing to respondents
Interval Scale
Numbers used to rank objects also represent equal increments of the attribute being measured
Differences can be compared
Relative score
Entire range of statistical operations can be employed for analysis
Ex. On a scale of 1 to 7, how would you rate the performance of natural gas as home heating...... (1 being least reliable and 7 being most reliable)-Negative number to allow for a neutral stance-1 being strongly disagree and 7 being strongly agree-Number itself is not meaningful, its where the customer is on the scale
Equally spaced buckets-Good option when asking for numbers that are hard to recall or sensitive questions (how many packs of gum did you buy in the last year / what is your income)
Ex. how satisfied are you with the economy right now? - very dissatisfied, dissatisfied, neutralm satisfied, very satisfiedex. What is your income?10,000-19,999 / 20,000 - 49,000
Ratio Scale
Type of interval scale with meaningful zero point
Possible to say how many times greater or smaller one object is than another
Only scale that permits comparisons of absolute magnitude
RAW data
What is your income ($)? / How old are you?Best data you can have - not being forced into buckets or forcing you to choose
Likert Scale
INTERVAL
respondents indicate degree to which they agree/disagree with a series of statements
series of statements where people agree or disagree
Include statements written in reverse (reverse code)Scales of 5 and 7 are good - neutral point
disagree strongly, disagree somewhat,neutral, agree somewhat, Agree strongly
Semantic Differential Scale
A unique bipolar ordinal scale format that captures a person's attitudes or feelings about a given object
respondents rate each attribute object on a number of five or seven-point rating scales bounded by polar adjectives or phrases
The midpoint is a neutral point
When coding in excel the notches would be coded:-2 , -1 , 0 , 1 , 2 → skews left or right based on answers not so much value of the numbers
Given example: This is a poorly worded scale because no one prefers high price or spotty quality
another ex.a Filters _ _ _ _ _ Post Stories
Viewing Stories _ _ _ _ _ Post stories
DM _ _ _ _ _ Group Message
Ways questions can be asked incorrectly
-Leading Questions:
-Double barreled questions
-vague questions
Leading (Loaded) Questions
Would you give me anything other than a 5 today? Was the food great?
Don't you think, bc its so greasy, fast food is one of the worst types of food?
double-barreled questions
two questions in one
-Are you satisfied with the price and the service of TacoBell?
Is the questions applicable to all respondents?
Why do you like fast food? - assuming that you eat/like it
3 forms of bias
-selection
- mortality effect
-testing (demand) effect
selection bias
who i'm selecting introduces bias
-Focusing on consumers at Neiman Marcus and trying to relate them to all of fashion valley
Mortality effect
data becomes more sparse overtime because people get tired of filling things out
-Not high quality data
-Can avoid this with incentives that build over time (complete 60%, you get this)
Testing effects (Demand effect):
how the question is asked induces people to give the ideal response
-2016 election, did not want to admit who they were going to vote for
variance
the spread between numbers in a data set
measure of population dispersion
How different are the individuals in the population
Subtract each person's score from the mean and then square it → as to not cancel anything out → then take into account sample size
standard deviation
a computed measure of how much scores vary around the mean score
hyptheses
statistical tests of relationships (changes in X or the effect of X on Y)
Ex. reducing package size will have a positive effect on sales
Restaurants allowing customers to use Groupon based deals will lead to an increase in sales
Introducing flavored beer will increase market share
Null Hypothesis (H0)
there is no difference between two groups
The effect of X on Y is non-existent
Alternate Hypothesis (Ha)
there is a difference between two groups
X indeed has an effect on Y
3 types of hypothesis testing
1. comparing mean to a benchmark
2. comparing two means to each other
3. comparing two proportions to each other
comparing mean to a benchmark
Useful benchmarks are the lowest side of the scale (0), the neutral point (4), or the high side of the scale (7) - obviously this is just when using a scale
comparing two means to each other
Within a given variable, comparing the averages of two groups
Ex. Loyalty (1-7) → avg loyalty of millennials vs avg loyalty of gen z
Willingness to recommend Netflix (1-7) → parents vs non parents
Avg between two nominal groups
comparing two proportions
Comparing two percentages
Comparing nominal variables (0,1 & 0,1)
In general, to test hypotheses, you need to
Calculate your standard error
How reliable your results are (reliability)
Calculate your test statistics (t-value) based on the standard error
Compare t-calc to the expected t-value (1.96 to have 95% confidence in your results, 2.58 to have 99% confidence)
If |t-calc| is greater than 1.96, then reject the null
proportion parameters
Mean: μ
Variance: sigma2
Std deviation: sigma
Standard err32qor: how reliable the results are
sigma / square root of N
sample statistics
Mean: X bar
0nmnmn/
Variance: s2
Std deviation: s
Standard error: how reliable the results are
S / square root of n
correlation language
Correlation always bounded between -1, 1
Population correlation (p)
Sample correlation (r)
R=0 → absence of linear association (non linear)
GDP (health of the economy)
Positively or negatively correlated with sales? → Positively
Price sensitivities? → negative (more money so sensitivity goes down)
Size shoes you wear? → zero correlation
Measure of Causation: Regression Analysis
Statistical technique that is used to relate two or more variables
Objective is to build a regression model relating the dependent variable to one or more independent variables
Spurious Correlations
Completely random
Divorce rate in maine & per capita consumption of margarine
Number of people who died by becoming tangled in their bedsheets & total revenue generated by skiing facilities
Age of Miss America & murders by steam, hot vapours, and hot objects
Mistaking correlation for causation
As ice cream sales increase, the rate of shark attacks sharply increase
IV: ice cream sales
DV: shark attacks
On hot days people eat more ice cream and spend more time in the ocean
How to get causation right?
Conduct a randomized study
Study population → random smokers vs non smokers
IV: whether you smoke or not
DV: health
Other sets by this creator
History 101
94 terms
MKTG 372 Exam 1
29 terms
MKTG 377 Exam 1
81 terms
BA4O5 Exam 1
70 terms
Other Quizlet sets
Test 2 chapters 5,6,7 Acct. Vocab
22 terms
Chapter 1 Terms
21 terms
Biochemistry - Block 3 - Fall 2014
243 terms
Exam 1 - Cell Bio
64 terms