marketing 340 final
what are the 4 examples of machine observers
1 traffic counters 2 The PeopleMeter (tries to match tv programs with demographic profiles) 3 scanner-based research (bar code) 4 physiological measurement devices
what is a population correlation
= p = rho = unknown
what is a sample correlation
= r = computed from a sample
correlation coefficients are always btwn what
-1 and 1
what are the 4 dimensions that classify observational research
1 human vs machine observation 2 natural vs contrived observation 3 open vs disguised observation 4 direct vs indirect observation
2 problems that can come with asking respondents questions
1 respondents make errors (memory errors or social desirability) 2 respondents cant answer
what are the 4 factors of experimental design. a set of procedures defining an experiment
1 the independent or predictor variables that are manipulated (x) 2 the subjects (ppl, stores ect) that participate in the experiment 3 a dependent variable that is measured (y) 4 a clever plan for dealing with extraneous factors - choose a good experiment design
in a brand awareness study, 25 of a group of 35 males identify the brand correctly and 15/35 females identify it correctly. the researcher wants to investigate whether the difference btwn men and women and their ability to identify the brand correctly or not when it comes to interpreting results from a cross tab, percentages can be computed in several diff ways. what is the correct % for the cell - "male identified brand correctly" a 62.5% b71.43% c35.71% d50% e57.14%
b 25/35 = 71.43%
what technique do you use if you have 2 interval/ ratio scaled variables
correlation coeff sample correlation = r population correlation = p - rho
when you wanna describe the relationship btwn 2 variables you complete the ?
correlation coefficient
what is a merger of 2 frequency tables
cross tab
what describes 2 nominal/ ordinal variables at the same time.
cross tabulation
which experimental design has the following symbols: experimental group: R O1 x O2 control group: R O3 04 a static group b posttest only control group design c one group pretest-posttest design d pretest posttest control group design
d
(t/f) when you start with analyzing data from a survey, it is best to start with bivariate analysis
false
laboratory experiments have (low/high) internal validity (low/high) external validity
high internal low external
what extraneous variable refers to specific events or factors that are external to the experiment but occur at the same time as the experiment.
history
a test statistic measures ..
how close the sample is to the null hypothesis
what asks "did you measure what u wanna measure"
internal validity
since a one shot design has no control or pretest, what kind of validity does it lack
internal validity and most likely external only good for suggesting causal hypothesis
what are the main drawbacks of physiological research
is the physiological change a valid measure for future sales or attitude change subjects are placed in artificial setting
experiments in marketing are either ... laboratory or field
laboratory = experiments conducted in a controlled setting field = experiments conducted outside a lab in an actual environment
correlation coefficient measures what
linear relationships only. nonlinear cant be measured with a linear correlation coeefficent
field experiments have (low/high) internal validity (low/high) external validity
low internal high external
when the pvalue is LARGER than the critical value ....
DO NOT reject the null
when the pvalue is LESS than the critical value ....
REJECT the null
what extraneous variable refers to changes in test units themselves that occur with the passage of time.
maturity ex: i am the same person but i give more effort during beginning of the study. not because of external distractions
measuring physiological reactions: what does a voice pitch analysis do
measures emotional reactions through physiological changes in a persons voice
measuring physiological reactions: what does a psychogalvanometer do
measures the galvanic skin response involuntary changes in the electrical resistance of the skin
natural vs contrived is ..
natural= researcher plays no role and those under observation should have no clue contrived= observation of an artificial situation (to speed up data collection, to control extraneous effects, to measure physiological reaction)
wheh X inc and Y tends to dec.. you have a ? correlation
negative
when higher values of x are associated with lower values of Y.. you have a ? correlation
negative
do you know the relationship btwn the variables in a cross tab
no
if knowing X doesnt help you know anything about Y.. you have a ? correlation
no correlation
what kind of data do you get from observational data
non verbal qualitative or quantitative past current or future
what does o represent
o = refers to the process of taking measurement on 'observing' the dependent variable
what is the systematic process of recording human behaviors of pattern of occurrences without normally communicating with the people involved
observational research
O1 X O2 is ..
one group pretest - posttest
what is a pre experimental design with a pre and post measurement but no control group
one group pretest - posttest
what design is used most frequently for testing changes in established products or marketing strategies
one group pretest - posttest the fact that the product was on the market before gives u O1 and post test is O2
X O1 is
one shot design
a pre experimental design with NO pretest observations, NO control group, and an after measurement only
one shot design
open vs disguised observation is..
open = people know theyre being watched disguised = people dont know theyre being watched
explain mystery shoppers (a type of human observer)
people who pose as customers and shop at a company's own store and its competitors purpose is to collect data about customer-employee interactions, compare prices, displays want to gather observational data about store
what extraneous variable is systematic differences between the test groups and control group due to a biased selection process
selection bias
what is it called if you have a strong positive correlation of 2 variables but they exist due to third variable
spurious correlation
in ? group extraneous variables are better controlled, but the 2 groups may be different before the experiment (lack of randomization - selection bias)
static group design
test units are NOT randomly assigned to treatment groups in what test
static group design
experimental group X O1 control O2 what is this
static group test
bivariate analysis
statistical investigation of the relationship of 2 variables
Explain ethnographic research (a type of human observer)
studies human behavior in its natural context goal: what makes people do what they do marketers often spend a lot of time with the customer involves informal interviews
guinea pig effect is
subjects change their normal behavior or attituted when they know theyre in an experiment
what is the correlation coefficient
summarizes the strength of linear association btwn 2 variables
(t/f) the correlation coefficient requires that data be measured on at least interval scales
t
what is correlation
the degree to which changes in one variable are associated with changes in another variable. allows us to infer traits about and predisposition of customers by asking just a few questions
measuring physiological reactions: what does a pupilometer do
the device observes and records changes in the diameter of the subjects pupils
(t/f) a sample of respondents is asked to taste 2 coffees (A or B) and choose the one they prefer. in order to investigate the difference between men and women and their choice between A or B coffee, you should use cross tab and a chi square test
true
(t/f) experimental research attempts to find causal relationships among variables
true
(t/f) lab experiments typically provide greater support for internal validity than external validity
true
what are machine observers
use a device to record aspects of behavior.
"install store layout A in store" = " measure satisfaction with respect to the several different dimensions in stores" =
x o
what is the independent variable
x the mover/ causer
what does x represent
x = indicated exposure to an experimental treatment
what is the dependent variable
y influenced/ moved
(t/f) in an observation study, we can observe behavior, physical personal characteristics, or cognitive phenomena
false
what are human observers
a marketing researcher (a person) observes people or phenomena
explain one-way mirrors (a type of human observer)
also can be cameras popular in focus groups
a marketing researcher downloads a data set from the internet about political attitudes. the sample has 100 respondents and had a likert scale. a researcher has info on what political party they are and what party they voted for. the researcher wants to investigate whether the people in this sample who voted which party in this election are equally likely to have a membership in that party which is true a this hypothesis can be tested with a ttest b for this study, the researcher doesnt have to worry about scale levels c a cross tab can be constructed to investigate d this is an ex of univariate e all of above
c
if the correlation coefficent between 2 variables is -.75 and the corresponding p-value from SPSS is 0.00 this means that a there is a weak positive relationship between the variables b there is a strong neg relationship between the variables c there is a weak neg relationship between the variables d there is a strong positive relationship between the variables
c
in a brand awareness study, 25 of a group of 35 males identify the brand correctly and 15/35 females identify it correctly. the researcher wants to investigate whether the difference btwn men and women and their ability to identify the brand correctly or not the Pvalie corresponding to the computed value of the chi-squared test statistic in the prev question is .018 the researcher had decided to use a sig level of 1% the researcher should a reject the null hypothesis b accept the null c not reject the null d do not reject the alternative e none
c do not reject
"when the occurrence of x inc the probability of the occurrence of y" = ?
causality
cons of observational methods
cognitive phenomena (motivations, attitudes, intentions, feelings) cant be observed may be contrived - observed behavior may be different than real world situation may be $$ and time consuming - observed bias possible (interpretation of data may be a prob and not all activity can be recorded) only public behavior is observed (possible invasion of privacy)
list the main pros of observational data:
communication with respondent isnt necessary (bc you get nonverbal data) -actual behavior doesnt lie (avoid social desirability) - no need to rely on respondents memory certain data may be obtained more quickly and accurately observation methods are often combined with -experimental methods to create contrived (Artifical) situations -surveys to provide supplemental evidence
what technique do you use if you have one nominal/ ordinal value and one interval/ ratio variable?
compare means and use a Z test to test for difference in those means for the population
a market researcher for sears investigates whether people with a college degree spend more money than people without a degree. Which of the following isnt true a the null hypothesis is H0: μ1 = μ2 b a pvalue can be computed from a test statistic to test H0 c the test statistic you use for this scenario is the t-test statistic d there are 3 variables involved here: people with a college degree, people without a college degree and amount of $ spent e all of the about
d varibales = t-test group h0: μ1 =μ2 p val comes from ttest
Using a stopwatch to determine the average waiting time for a customer at a drive-through location at McDonald's is an example of what type of observation?
direct observation
direct vs indirect observation is..
direct= observe the actual behavior indirect= infer behavior through some record of past behavior (aka physical trace evidence)
what are testing effects
effects that are a by product of the research process ex - milgram experiment
what are the 3 examples of human observers (list them)
ethnographic mystery shoppers one way mirror
what is a research approach in which one or more variable manipulated (x) and the effect on another variable (y) is observed
experiment
what asks " can you extend your results to the real world"
external validity
what are all variables other than X and Y
extraneous variables
(t/f) correlation always means that a causal relationship exists between 2 variables
false
(t/f) eye tracking monitors are only effective in measuring ad effectiveness
false
(t/f) if the value of r = -1.0 there is a perfect positive relationship between the 2 variables under study.
false
(t/f) in an experiment, the researcher manipulates the dependent variable and measures the effect on the independent variable
false
positive correlation = no correlation= negative correlation=
positive correlation = correlation larger than 0 no correlation= correlation close to 0 negative correlation= correlation less than 0
experimental: R X O1 control: R X O2
post test only control group
what is an experiment design that has Random Assignment of subjects to experimental and control groups but no premeasurement of the dependent variable
post test only group design
what is the diff btwn post test only group design and pretest post test
post test only group design is shorter. offers time and cost advantages ppl are less likely to drop out cant always get a pre test
what are 2 designs that are better in controlling extraneous variables and why
pretest - posttest control group post test only control group bc a control group is used and subjects are randomly assigned to treatment groups
what is an experiment design that involves Random Assignment of subjects or test units to experimental and control groups and post measurements of both groups
pretest post test
experimental: R O1 X O2 (O2-O1) = ? control : R O3 O4 (O4-O3) = ? IS..
pretest post test control group
what are studies in which the researcher lacks complete control over the scheduling of treatments or must assign respondents to treatments in a nonrandom manner
quasi experiments
when extraneous variables are not adequately controlled for and no randomization is employed..
quasi experiments
What does R represent?
random assignment of tests units to treatments
what are possible solutions to deal with selection bias
randomization or matching
measuring physiological reactions eye tracking monitors
records how the subject actually reads or views a web page, ad, ect measures unconscious eye movements
what is external validity
refers to whether the cause and effect rltnship found in the experiment can be generalized
internal validity is
refers to whether the manpulation of the indep variables or treatments actually caused the observed effect on the dependent variables
human vs contrived observation is useful for..
routine repetitive behavior for hard to observe physiological changes may be more intrusive/ expensive than human observation
