Term
|
Definition
-assignment of numbers to responses based on guidelines.
-easy to quantify, we can manipulate numbers.
-we can summarize quantified responses freom a large sampe more than non-quatified
- |
|
|
Term
|
Definition
1)numbers follow a rank order
2)we can compare intervals or differences between pairs of numbers
3)divide one number by another to create a ratio |
|
|
Term
|
Definition
| categorical numbers are numerals used for identification only, value is irrelevant |
|
|
Term
|
Definition
| the numbers possess a rank order; allows us to look at the median |
|
|
Term
|
Definition
| all properties of an ordinal scale, difference is between scale value (like speed limit) |
|
|
Term
|
Definition
| all properties of interval scale, ratios have meaningful interpretation (most versatile scale of all) |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
| Classes of variable reason (4) |
|
Definition
Attributes
Behavior
Beliefs
Attitude |
|
|
Term
|
Definition
| personal or demographic characteristic (age, size of house or children) |
|
|
Term
|
Definition
| related to how often do you go to the store, what magazine do you read |
|
|
Term
|
Definition
| knowledge and what respondents consider being true, whether or not it is true |
|
|
Term
|
Definition
very critical, impacts our decision making and reflects judgements (similar to beliefs)
Cognitice, affective, conative --> key determinant of behavior |
|
|
Term
| reporting scaling questions |
|
Definition
| scaling questions to allow the respondents to provide an opinion or belief level, and allow the researchers to perform analysis on those |
|
|
Term
| Scales allow us to (3 things) |
|
Definition
1) to observe certain behavior
2) people to react to partially structered stimuli
3)to evaluate |
|
|
Term
|
Definition
|
|
Term
| 2 levels of response categories |
|
Definition
even - specific choice either good or bad
bad - allows neutral choice |
|
|
Term
|
Definition
| process of examining the completed data collection forms and taking whatever corrective action is needed to ensure that the data is of high quality (collected correctly) |
|
|
Term
| Preliminary or field audit (3) |
|
Definition
to ensure we have readable, legible forms
1- inappropriate responses
2- incomplete interviews
3- forms not readible |
|
|
Term
|
Definition
| things that the gov't thinks can go wrong, WILL go wrong! |
|
|
Term
|
Definition
1) transforming responses into meaningful categories
2) assigning numerical codes
3) facilitate numerical analysis |
|
|
Term
| preliminary data analysis |
|
Definition
identify measures of central tendency
--> mode, median, mean |
|
|
Term
| Measures of dispersion RV(2) |
|
Definition
range - difference between smallest and largest, this is the easiest to commute
variance - measure of dispersion of respondents around the mean |
|
|
Term
| frequency distribution 1,2 (2) |
|
Definition
one way table - shows distribution of single table
2 way table - shows 2 different variable |
|
|
Term
|
Definition
| consists of computing measures of central tendency + dispersion as well as constructing tables and research summaries of nature of study variables |
|
|
Term
|
Definition
null hypothesis - the original hypothesis
alternative hypothesis - they compliment each other, are mutually exclusive and collectively exhausted
-use alternative if null is rejected |
|
|
Term
| type 1 error (hypothesis testing) |
|
Definition
| committed if null hypothesis is true but is rejected |
|
|
Term
|
Definition
| committed if null hypothesis is false but is not rejected |
|
|
Term
|
Definition
| guideline that specifies sample evidence necessary to reject null hypothesis |
|
|
Term
| 7 Steps of hypothesis testing |
|
Definition
1) Set up Null Hypothesis first and then set up the alt hypothesis.
2) identify nature of sampling distribution curve and specify the appropriate test statistic (what are we calculating)
3) determine whether hypothesis is 1 tail or 2 tail?
4)take into account the specified significance level which can be found on (appropriate statistical table) and determine the critical value
5) state decision rule for rejecting the null hypothesis
6)compute the value for the test statistic from sample data
7) using decision rule (step 5), either reject or accept the Null Hypothesis |
|
|
Term
| 2 factors for choosing appropriate analysis procedures |
|
Definition
1) # of variables that you are analyzing, univariate (one variable (gender) or multivariate) 2 or more
2) nature of the data we are collecting
- data nominal? ordinal? interval? or ratio? |
|
|
Term
|
Definition
| table that looks at 2 variables (chi square or two way table) |
|
|
Term
|
Definition
| 2 independent samples and compare them statistically. |
|
|
Term
|
Definition
| degree and direction of association as a single number |
|
|
Term
| spearman correlation coefficient (1) |
|
Definition
| linear between 2 ordinal rank order variables (identical) |
|
|
Term
| pearson correlation coefficient (2) |
|
Definition
| interval/ratio greater than 1 variable |
|
|
Term
|
Definition
generate a mathematical function or equation linking variables.
-single regression analysis
- multiple regression analysis |
|
|
Term
|
Definition
| using trends from the past to predict future |
|
|
Term
|
Definition
| using modeling of various stats that allows to observe relationship overtime |
|
|
Term
|
Definition
| generates mathematical relationship regression between one variable designated as the dependent (x) and the other independent (y) |
|
|
Term
| multiple (multi- collinearity) regresson anal |
|
Definition
| when there are 2 or more independent variables, highly correlated |
|
|
Term
|
Definition
| 1 variable is designated as a dependent variable and all the rest are independent variable |
|
|
Term
|
Definition
| there are no dependent or interdependent designations - all treated equally - looking for underlying relationships |
|
|
Term
| ANOVA - analysis of variables |
|
Definition
technique for detecting relationships of interval and ratio and 1 or more non metric variables
high level, mid level, low level |
|
|
Term
|
Definition
| said to be present when the relative impact of treatment levels of one factor varies across treatments levels of another factor |
|
|
Term
interdependence techniques (2)
(factor analysis) (1) |
|
Definition
| - fa- data + variable reduction techniques that attempts to partition into a given set of variables into groups (factors) of **maximum correlated variables** |
|
|
Term
|
Definition
| segments objects into groups so that members of those groups share as much info as possible |
|
|
Term
| multi dimensional scaling - |
|
Definition
intended to find the underlying dimension, multiple factors
*used to identify relative positions* of brands as perceived by customers. Perception rather than fact |
|
|
Term
|
Definition
| technique for deriving utility values that customers attach to different levels of an objects attributes |
|
|
