buad 302 exam 1

why marketing managers shouldn't rely on intuition to make decisions

1. people have a hard time learning from experiences

2.people are bad at using intuition for dollar value of information

3. People are limited and biased information processors

-availability bias

ex. do more words start with a K or k is the third letter

4. better than average effect

-93% of americans see themselves as better than the median

why marketers make decisions

1. practice marketing

2. implement marketing concepts

3. select and execute marketing strategy

marketing research

data --> information --> action

process of designing, gathering, analyzing, and reporting info that may be used to solve a specific marketing problem

market research

systematic gathering, recording and analyzing of data with respect to a particular market, where market refers to a specific group in a specific geo region.

role and value of marketing research

relies heavily on social sciences for theory and methods

Methods

-are diverse

-span variety on qualitative and qualitative techniques

-borrow from disciplines of psychology, sociology, econ, anther

Descriptive research

gathering statement/ facts

ex. demographics of jeep users

finding out ppl satisfaction towards jeeps

explanatory/diagnostic research

explaining marketing mix actions or data

ex. decide whether the correct market is being targeted for the product

explain trends found in satisfaction

predictive research

predicting results of a planned marketing decision

ex. new marketing plan for diesel fuel how will demand increase for diesel cars

applied research

conducted when a decision must be made about a specific problem

basic research

attempts to expand limits of knowledge rather than solving pragmatic problem

-used to better understand the world we live in

product

-development and introduction for new product

-branding (logo, name)

-positioning

-where they fall on perceptual map

perceptual map

positioning map

-picture the relative position of products on two or more product dimensions important to consumers purchase decisions

place

-decisions on locations, channels, distribution partners

-Retailing research

retailing research

focus on trade area analysis, store image/perception, in store traffic patterns, location analysis

behavioral targeting

placing ads based on users previous surfing history

shopper marketing

marketing to consumers based on research of the entire process consumers go through when making a purchase

Promotion

influence on any companys sales

-essential that companies know how to obtain good returns from promotional budgets

ex. ad schedule

IMC integrated marketing communications

-advertising effectiveness

-attiduinal research

-sales tracking

Price

decisions involve

-pricing new products

-establishing price levels in test marketing

-modifying prices of existing products

conjoint analysis

one of most widely used qualitative methods

-measures preferences for product features, to learn how changes to price affect demand for products or services

-helps to forecast likely acceptance of a product being brought to the market

-can ask directly or as a trade off among alternatives

Segmentation studies

creating customer profiles

-benefit and lifestyle studies: examine similarities and differences in consumer needs

-researchers use studies to identify two or more segments within the market for particular companies products

Ethnographic research

used in segmentation to study consumer behavior as activities embedded in a cultural context and laden with identity

-requires extended observation of consumers in context

ethical questions

-unethical pricing

-unnecessary or unwarranted research services

-client confidentiality issues

-use of branded "black box" methodologies: offered by research firms that are branded

-do not provide info about how their methods work

marketing is growing

new technology

-new data collection tools, ex twitter, gps

-new tools and techniques make more choices to choosing right technique

5 skills for marketing research firms

1. ability to understand and interpret secondary data

2. presentation skills

3. foreign language competency

4. negotiation skills

5. computer proficiency

measurement

assigning numbers or other symbols to characteristics of objects according to certain prespecified rules

-the rules for correspondence should be standardized and uniform

-rules must not change overtime

one to one correspondence

btw numbers and characteristics being measured

numbers

usually assigned for one of two reasons

-statistical analysis

-facilitate communication rules

measurement

an integrative process of determining the intensity (amt) of information about constructs, concepts, or objects

1. construct selection/development

2. scale measurement/implementation

construct

-an unobservable concept made up of a set of component responses or behavior that are thought to be related

-an abstract concept that represents real world phenomena

ex. age, gender- easy

brand loyalty, attitude - hard

operational definition

precise definition that specifies the activities or operations to measure the concept

easy- age, gender

hard- brand loyalty

scale measurement

process of assigning numbers or label to persons, rules, objects, or events

ex, strongly agree, disagree

description

you can assign objects to categories

order (magnitude)

you can order objects in terms of having more or less of some quality

distance (equal intervals)

distance btw two adjacent points

origin (abs zero)

zero means something

nominal scales

a scale in which a numerical value merely identifies a level

-ssn

-zip code

-gender

-color

-no mathematical properties. one cannot compare them

-categories need to be mutually exclusive and exhaustive

use bar graphs, frequency and pie

ordinal scales

scale in which the number is not only identity the level but also indicate some order between levels

-level of service : low, medium, high

-preference for brands- rank order

-mild, medium, hot

difference btw numbers not meaningful

stats: mode, median

rank order scales

paired comparisons

Q sort

itemized category

interval scale

scale in which the difference btw levels is so quantified. there is no abs zero. equal distance btw scale values.

-shows absolute differences btw each scale pt

possible stats

-frequency tables, mode, median, mean, standard dev, t tests, regressions, ANOVA, factor, cluster, some conjoint

interval scale: ordinal scale + meaningful differences

ex. 1-10 agree-disagrees

attitude measurement

semantic differential

LIKERT

continuous (ex.

ratio scale

abs zero exists

-weight

-income

-length

ex. sales on a given day

constant sum

Likert Scales

Interval

most popular way to obtain attitudes

usually five or seven point rating scale

advantage

-researcher can obtain summated data across items to tell attitude

disadvantage

-tricky to come up with statement

-takes respondents longer to read

reliablilty

extent to which scale produces consistent results if repeated measures are made (reproducibility)

-is is free of random error

validity

extent that the scale measures what it intends to measure , free of systematic error

test-retest reliability

repeat same survey two weeks later to the same sample of people (should get same results)

equivalent form

use similar questionnaires then compare the correlations of a similar test items

internal consistency

use same instrument with two different sample, or sometimes split the sample into halves

scale validity

researcher makes a judgment as to whether the measure "looks like" it measures what its suppose to

ex. income vs disposable income

content validity

did the scale provide adequate coverage of the topic

discriminatory power

the scales ability to discriminate between categorical scale responses (points)

negatively worded statements

forced or non forced

no natural option - forced

neural - non forced

balanced vs unbalanced scales

not equal for each, ex. more negative choices than positive

non comparative scales

scale format that requires judgement without reference to another object, person, concept

ex. usage behavior, smily face

comparative scales

scale format that requires a judgment comparison with an object, person or concept

ex. constant sum scale, rank order scale

IAT reaction time

subject categorize words with subjects

ex. neg pos with pic of person

shows attitudes towards people

scale points

designated degree of intensity assigned to the responses in a given questioning or observation method

interval vs ordinal

interval scale is ordinal + meaningful differences btw responses

interval every point is equal distance apart

convergent validity

multi-item scales and represents a situation where multiple items measuring the same construct have a high variance over 50%

discriminate validity

extent to which a single construct differs from other constructs creating a unique construct.

single item scale

scale that involves collecting data about only one attribute of the object or construct being investigated

-ex. age scale

multi-item scale

a scale format that simultaneously collects data on several attributes of an object or construct

-ex. collecting attitudinal, emotional, behavioral data use multi-item scale

-each statement has a rating scale attached to it and the researcher often will sum the ratings on the individual states to get an overall rating for the construct

data preparation

converting information for surveys and other data sources so it can be used for statistical analysis

-essential for converting raw data into usable coded data for data analysis

Data preparation and analysis

data prep

validation

editing and coding

data entry

data tabulation

data analysis

data validation

process of determining, to the extent possible, whether a survey's interviews or observations were conducted correctly and are free of fraud or bias

-most tedious step

process covers

-fraud

-screening interviewers

-procedure

-completeness

-courtesy

curbstoning

falsifying data

filling out the survey on own to lie

-often "call backs" are used to verify results of surveys

editing

process where raw data are checked for mistakes made by either interviewer or respondent

-Did the interviewer ask the proper questions?

-accurate reading of open-ended questions?

-accurate recording of answers?

-correct screening questions?

coding

grouping and assigning values to responses from surveys

-numerical 0-9

-open ended questions are harder to code

4 step coding process

list potential responses

responses are assigned values within a range

used for open ended to give value to answers

assign coded value, must be done manually

data entry

tasks involved with the direct input of the coded data into some specified computer software package that ultimately allows the research analysts to manipulate and transform the raw data in to useful info

-makes sure data does not have any errors or inconsistencies

ex. missing answers

incorrect codes

error detection

identifies errors from data entry or other sources

-usually better to avoid errors at question design stage

-can program to show outliers

missing data

a situation where the respondent does not provide an answer to a question

-most online surveys require to answer all questions for this reason

-when there is missing data

-can replace with response of a similar candidate

-look at answer to similar questions

-use the mean of the subsample of respondents with similar characteristics

-use the mean of entire sample

ONLY impute data if it is missing at random, if not it will cause trouble

Data Tabulation

process of counting the number of observations that are classified into certain categories

descriptive statistics

one way tabulation

categorization of single variables existing in a study

-shows the frequency count who gave each possible answer to each question on questionnaire

-used to locate missing data

-frequency table

cross-tabulation

simultaneously treating two or more variables int he study; categorizing the number of respondents who have answered two or more questions consecutively

ex. number of male and females who spent more than $7 at mcdonalds

graphical summaries of data

frequency chart

histogram

bar graph

line

pie

histogram

interval, ratio

frequency, bar, pie

nominal, ordinal

when to code

When testing a hypothesis (deductive), categories and codes can be developed BEFORE data is collected.

(when designing questionnaire)

When generating a theory (inductive), categories and codes are generated AFTER examining the collected data.

measures of central tendency

mean, median, mode

mean

average value within distribution

measures central tendency

insensitive to extreme values

-best for interval or ratio

median

middle number

-if even number then it is the average of two middle

-best for ordinal

uses stem and leaf plot

find middle 10's then middle of that number

-more robust to help find outliers

mode

number that appears most often

-best for nominal

range

spread of data

identifies the end points of value distribution

largest value- smallest value

ex. largest: 5 smallest:1 range :1

standard deviation

average distance of the distribution values of the mean

variance

average squared std deviation

ex. std dev squared

measures of dispersion

range

std dev

variance

descriptive statistics

make data usable from survey

not about whole populations

inferential statistics

generalize results to a population

null hypothesis

what statisticians test

if the null hypothesis is accepted then reject null and nothing changes

if rejected and alternative accepted then need to make change

alternative hypothesis

opposite of the null hypothesis

if alternative hypothesis is correct then reject null and a change in behavior needs to be made

sample

subset of population

population

all the people who share common characteristics

population parameters

Summary description of a variable of the population

Greek lower-case letters; e.g. µ : mean or σ - std dev

sample statistics

A summary description of a variable of the sample, hence is computed from data

The sample statistic is used as an estimate for the population parameter

English letters for notation, e.g. or s x bar

nominal ordinal

Counts, frequency tables, proportions (percentages), mode

ratio interval

mean, median, variance, std deviation, range

dispersion

frequency: nominal, ordinal, interval, ration

range: ordinal, interval, ratio

std dev: interval, ratio

central tendency

mode: nominal, ordinal, interval, ratio

median: ordinal, interval, ratio

mean: interval, ratio

histogram peak

frequency

under peak = population

management decisions

made for the population

based on the smooth line

Population distribution

frequency distribution (histogram) of the population, usually a smooth line, but is

unknown

for height it seems to be a normal curve for income it seems to be a skewed curve (how to tell?)