Term
Independent variables are MANIPULATED in some systematic way |
|
Definition
|
|
Term
The dependent variable is used to evaluate the influence of the independent variable. |
|
Definition
|
|
Term
Experiments go beyond the "what" of a problem in order to arrive at the.... |
|
Definition
|
|
Term
3 pieces of evidence are needed to support the assumption that variable A causes variable B. They are... |
|
Definition
1. Correlation 2. Sequence (variable B must appear after variable A has appeared, or must change after variable A changes) 3. Absence of other causal factors |
|
|
Term
|
Definition
Variable B must change/appear after variable A changes. |
|
|
Term
|
Definition
the extent that one can eliminate alternative explanation for the observed experimental results |
|
|
Term
Internal Validity Threats |
|
Definition
*History *Maturation *Testing Effect *Instrumentation *Mortality *Selection Bias |
|
|
Term
Requirements for Causality |
|
Definition
1. Events must take place in the proper order (( to determine if changing the amount of water affects the firmness, crispness and color of pizza dough, you have to change the water first) 2. events must take place at the same time---con-comitant variation-causes and effects must occur or vary together (crust quality and water manipulation varied together) 3. Alternative explanations must be reduced: affects an experiments internal validity |
|
|
Term
Problems affecting internal validity (confidence that your experiment has occured bc u manipulated something to get a particular effect): INTERACTION |
|
Definition
Interaction occurs when the INDEPENDENT variable is more likely to be noticed and reacted to then it would be without the initial measurement (like a survey) |
|
|
Term
The threat of PREMEASUREMENT to internal validity occurs... |
|
Definition
WITHOUT exposure to the independent variable. All observed attitudinal or behavioral changes are the result of exposure to the initial measurement instrument |
|
|
Term
Problems affecting internal validity: HISTORY |
|
Definition
any events or influences beyond those intentionally manipulated by the researcher that occur during the experiment and that have the potential to affect the experimental outcome as measured by the dependent variable(s)---ex: studies of a soft drink, but there is unusually warm or cold weather prior to the study |
|
|
Term
Problems affecting internal validity: Maturation |
|
Definition
how a respondent's attitudes, behaviors, and physiology change during an experiment. ((By the end of the experiment, they are sick of being there, so their answers might not reflect their true feelings)) |
|
|
Term
Problems affecting internal validity: Instrumentation |
|
Definition
changes made to the measurement instrument (ex: questionnaire) or data recording techniques during the experiment. |
|
|
Term
Problems affecting internal validity: Selection |
|
Definition
occur whenever the characteristics of the control and the test group are not equivalent before the experiment. They differ in relevant demographics, attitudes, behaviors, their initial level with regard to the independent variable, or their likelihood to respond to the indep. variable. These selection threats can be reduced by using RANDOM SELECTION and ASSIGNMENT OF INDIVIDUALS TO THE TEST AND CONTROL GROUPS |
|
|
Term
Threats to internal validity: Mortality |
|
Definition
typically arise when different types of individuals from each group drop out of the study..
*If 10% of the participants drop out of the control group (but they are aged 18-24) and 10% drop out of the other group (but they are aged older), this affects the mean ages of each group and thus their responses to the dependent measures. |
|
|
Term
Instrumentation (further explained) chart 7.1 *Changes in the measurement instrument *changes or noncomparability in recording techniques *changes in data collection methods *inconsistent interviewing |
|
Definition
**Changes in the measurement instrument: A particular survey question during the pretest to the experiment is confusing, so it is changed for the posttest, thus making it impossible to accurately compare attitudes on the pretest with the posttest.
**Changes or noncomparability in recording techniques ---using two different interviewers to record reactions, when the process requires judgement that could vary depending on the person.
**changes in data collection methods:using personal interviews and then a mail survey prevents the bank from confidently comparing attitudes in 1995 and 1996
**inconsistent interviewing: interviewing a group of people and spending a lot of time clarifying and probing for questions, but then the 2nd time around, minimizing those activities--can have a dramatic effect on the content |
|
|
Term
|
Definition
helps to contol the seven threats to internal validity |
|
|
Term
A researcher must identify the experimental design that provides the best balance between ACCURACY and COST. |
|
Definition
|
|
Term
True Experimental Designs |
|
Definition
*simulated pretest-posttest
*posttest-only with control *pretest-posttest with control *Solomon four-group design |
|
|
Term
true experimental designs: Simulared Pretest-Posttest |
|
Definition
*uses one group of randomly assigned respondents for the pretest measurement and one group of randomly assigned respondents for a posttest measurement control for premeasurement and interactional threats to internal validity, particularly in experiments dealing with consumer attitudes and knowledge. |
|
|
Term
true experimental designs: Simulared Pretest-Posttest |
|
Definition
*uses one group of randomly assigned respondents for the pretest measurement and one group of randomly assigned respondents for a posttest measurement control for premeasurement and interactional threats to internal validity, particularly in experiments dealing with consumer attitudes and knowledge. ---> design assumes that the control and test groups are equivalent |
|
|
Term
Posttest only with Contro |
|
Definition
utilizes 2 groups of respondents; cost roughly equic to simulated pretest-posttest; measures treatment effects by comparing postmeasures> one postmeasure obtained from the treatment group and one obtained from the control group |
|
|
Term
The posttest only with the control design is a more powerful design than the simulated pretest-posttest BECAUSE |
|
Definition
it controls for a greater number of threats to internal validity.
So, in summary, having a control group and an experimental group and measuring the results after the experiment controls the threats to internal validity and is a MORE POWERFUL DESIGN than the simulated pretest and posttest in which two seperate groups are chosen and their results are compared (pretest for group 1 compared to posttest of group 2) |
|
|
Term
The Pretest-Posttest with Control |
|
Definition
2 groups;
GROUP 1: Pretest-> Treatment -> Posttest GROUP 2: Pretest-> no treatment-> posttest
This design is appropriate whenever a researcher needs explicit evidence of a group equivalency before the start of the treatment or whenever there is some doubt to the extent of the group equivalence. |
|
|
Term
The Soloman 4-group design
FOUR GROUPS-SIX MEASURES |
|
Definition
*most powerful and most resource intensive experimental design
*controls for all 7 threats for internal validity |
|
|
Term
Four Groups, Six measures: The Soloman Four-Group Design |
|
Definition
Group 1: Pretest-> Experiment-> Posttest Group 2: Pretest--> XXXXXXX--> Posttest Group 3:XXX-> Treatment-> Posttest Group 4:XXX->XXXX-> Posttest |
|
|
Term
The Soloman 4 group design--4 groups, 6 measurements |
|
Definition
Groups 1 and 3 would receive the experiment 1 and 2 get the pretest, all get the posttest |
|
|
Term
|
Definition
experimental procedure that simultaneously measures the effect of two or more independent variables, each with different levels, on one or more dependent variables |
|
|
Term
Factorial designs consist of MAIN EFFECTS and INTERACTIONS |
|
Definition
Factorial Designs consist of MAIN EFFECTS and INTERACTIONS. |
|
|
Term
|
Definition
the separate influence of each independent variable on the dependent variables. Spokespersons (celebrity or not) and tone (humorous and serious) to see the affect. |
|
|
Term
interaction occurs when... (pg 152) |
|
Definition
the effect of the COMBINATIONS of main effects on the dependent variables
When the simultaneous effect of two or more independent variables is different from the sum of their independent effects. |
|
|
Term
|
Definition
If responses to a particular tone and spokesperson combo were different than they were with the main effects seperate. |
|
|
Term
1. Identify concept of interest 2.develop conceptual definition 3. develop operational definition |
|
Definition
Stage 1 of the measurement process |
|
|
Term
1. Identify appropriate level of measurement 2. Identify appropriate question type |
|
Definition
stage 2 of the measurement process: SPECIFY AN OBSERVABLE EVENT ((actual research question or data collection instrument)) is created |
|
|
Term
4 levels of measurement: NOMINAL ORDINAL INTERVAL RATIO |
|
Definition
|
|
Term
|
Definition
*very common * occurs when the goal is only classification of the measured characteristic or attribute. |
|
|
Term
|
Definition
yields consistent results over time hitting the same spot on the target board, even if its not the center |
|
|
Term
|
Definition
Hitting the target dead-on; measures what it is intended to measure |
|
|
Term
A PILOT STUDY is conducted to assess |
|
Definition
the RELIABILITY and VALIDITY of untest measures |
|
|
Term
|
Definition
abstract, unobservable concepts are linked to observable events |
|
|
Term
The Measurement Process (3 stages) |
|
Definition
STAGE 1: Refinement ---Identify and define concept of interest -develop a conceptual definition (what is an advertising attitude) --develop an operational definition (the guidelines used to evaluate the conceptual definition..advertising attitudes: happy vs sad, etc) STAGE 2: Observable Events --identify an appropriate level of measurement: Nominal, Ordinal, Interval, Ratio --identify an appropriate question type: open- ended vs. closed-ended
STAGE 3:Evaluation and Revision --Write (or rewrite) the question |
|
|
Term
|
Definition
simplest type of measurement for nominal questions, used to classify, usually YES OR NO questions |
|
|
Term
nominal level questions should be both |
|
Definition
Exhaustive (include every option possible) and exclusive (segment people into distinctive categories) |
|
|
Term
|
Definition
|
|
Term
procedures to use to differ ways of asking multiple choice questions |
|
Definition
Use a check-procedure (start in the middle of a list of answer choices) Use a split-ballot procedure--if you want to survey 300 people, split the survey up 3 ways with the choices in a varied order)
Checklist--check as many that apply. Should be in alphabetical order and contain equal numbers of positive and negative items. |
|
|
Term
Interval questions: 2 types of composition |
|
Definition
Rating Scales Noncomparative Rating Scales (ask respondents without an explicit frame of reference) example: How believablw or unbelievable do you think... |
|
|
Term
|
Definition
7 scale questions: The celebrity was Attractive..1 2 3 4 5 6 7.......Unattractive |
|
|
Term
|
Definition
simplified versions of the semantic differential, standard 11 rating scale from -5 to +5 for one term, such as trustworthiness |
|
|
Term
|
Definition
respondent indicates degree of agreement or disagreement with a series of statements: strongly agree, slightly agree, etc. |
|
|
Term
Ratio Level Questions: Constant Sum Scale |
|
Definition
Usually out of 10 or 100.. participant is asked to divide the points among the attributes |
|
|
Term
|
Definition
provides an index of the degree to which scores deviate from or are at variance with the MEAN |
|
|
Term
The square root of variance is the.... |
|
Definition
|
|
Term
If there is a high variance, it means the data is widely spread; varied. |
|
Definition
|
|
Term
A small standard deviation means |
|
Definition
there was a consensus of responses |
|
|
Term
|
Definition
|
|
Term
2 instances in which the median is preferred to the mean |
|
Definition
1. one or more extreme values in the data set 2. open-ended categories include, income data "over 100,000" has no end point, making it difficult to find the mean |
|
|
Term
Nominal and Ordinal use cross-tabulation, while interval and ration use correlation |
|
Definition
|
|
Term
|
Definition
change with every product class and describe a specific consumer-product relationship; provide info for message and positioning strategies |
|
|
Term
|
Definition
they do not change across product classes;provide info for media selection |
|
|
Term
|
Definition
Innovators (care about image, open to new products, shop for luxury) Strugglers (12% of adult population, health issues, least educated) Thinkers (knowledgeable about the world Believers (politically conservative A Strivers-- pay attention to trends, care what others think about them Experiencers-like sports, excitement, enthusiastic about their career Makers (practical, shop for comfort) |
|
|
Term
What's the difference btwn. DESCRIPTIVE and INFERENTIAL statistics? |
|
Definition
|
|