Market Research at Universität Münster

if the sample is equal to the population 

Experiment in small section, with the goal of prediciting sales of some proposed marketing actions 

1. Simulated test market if promising next step 2. standard test market (for example Bremen, isolated city, average city)

if this promising too 3 national rollout (very costy) 

1 use simple words and short sentences 

2 avoid ambiguity (mehrdeutigkeit)

3. avoid leading questions

4. avoid implicit assumptions

5. Avoid generalizations

6. avoid double barreled questions (doppelte Fragen)

7. avoid negations

generate insights to concepts (generating)--> investigate what kinds of needs/ wants customors have

and what challanges of problems they face

concepts to products (developing)--> develope an idea to solve this problems 

and launch it

product to money (monitoring)-> is it successful

earn money (why/ why not)

group of units about which to make judgments 

important and not trivial decison

size of population can differ 

for example people living in germany 

or students of economics 

random selection from population

each element has the same chance of being selected, 

every combination of elements is possible 

+ easy to realize 

only works well if we have a homogenous sampling frame and people do not differ much

customers-> which is the target segment buying behavior etc

product-> does the company offer an adequate product? quality standards, features etc

channels-> the product promoted within the right channels

competiors-> which exists and which products do they offer 

market share etc

trends/enviroment-> do enviromental influences exist? change in lifestyle etc

introduction

aim of study

ice breaking introduction question

crucial questions

control questions

socio dempgraphic data

thank you words

-

description of scale points-> exhaustive (ausführlich)

                                                    polar only very good very bad in between you can chose

uncertainity and ignorance of the respndent-> foreced choice (no neutral point)

allow a neutral ppint

allowance of a dont know categorie

balance of scales points-> balanced (same points in both direction from average)

unbalanced (unbalanced points for both directions)

anchoring the response-> with comparison to other 

without comparison

primary: how can we boost the sales of our product in grocery stores (quite specific question)

secondary: how many potential customers could we target with our new product

(already collected)

variation->  detect variation in your data that can serve as warning system (question is not measuring the intended construct) 

meaning-> can make sure your respndent understand the same you do (include open end question in pre test)

task difficulty (questions more accessible not year weeks for example better

respndents interest and attention-> include attention checks 

not the same strucutre question, do not repeat etc