Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Market Research Process Clarity
#1
Hello to all....I'm a new and aspiring Market Researcher. I'm learning things and have taken classes/courses on Research Rockstar. 


I have some questions concerning a particular phase of a survey research project.....

As I have learned, there is a process or sequence of events that happens when you're doing research for a client. The following is what I've been taught through the Market Research 101 course on Research Rockstar:

1. Kickoff meeting. The kickoff meeting consists of:

Phase 2a. Schedule
Phase 2b. Project Management Plan
Phase 2c. Sampling Plan
Phase 2d. Data Analysis Plan

Phase 3. Questionnaire Design 

Phase 4. Questionnaire Approval 

Phase 5. Programming and testing

Phase 6. Pre-testing

Phase 7. Data Collection (During some downtime in this phase, the deliverables plan is developed and updates are made to the data analysis plan).

Phase 8. Analysis, Charts, Synthesis

Phase 9. Deliverables: Report, Top line Montage, Transcript, Presentations,etc.


At the Programming and testing stage are what my question are about:

Q1. Can someone explain to me exactly what the programming part means?
Q2. What is the Market Research Analyst supposed to be programming?
Q3. And is the testing activity in step 5 the same or different than pre - testing activity in phase 6? If so how?
Reply
#2
Programming is the 'scripting' of the questionnaire into a survey system it so it will work on the web or by phone for collecting data. However, even pure paper-based questionnaires need to be turned into something that can be used to capture and process the data for analysis.

In pre-internet times, the research analyst wrote the questionnaire on paper, and a DP (data processor) expert converted it be used on the computer.

And where the task is split between the research analyst and someone from DP, then the analyst needs to check the DP scripting matches with the paper version of the questionnaire - eg spelling, layout, question types. The research analyst scripting themselves means this step shouldn't be required.

However, now, I'm surprised questionnaire design and programming are listed as separate tasks. Many people go straight to a survey program for questionnaire design and the DP role has merged with that of the Research Analyst, so the questionnaire is developed on the survey system right from the start of the process.

Pre-flight testing is still required though, to pre-validate the answers and ensure data is coming in as required/expected, to check error handling and edge cases are adequately checked, and to double check filter and routing conditions handle properly (eg check someone who doesn't read newspapers is skipped so as not to be asked further questions about newspaper readership), or that items like video, images etc display properly across different devices. It's easy to miss this and then end up wasting 1000 people's time and answers because of a design flaw.

A good survey application will not only include design of the questionnaire, but also fieldwork monitoring (eg quota management, screening, reminders and for phone work, appointment setting and diallers, interim reports) and then provide automation and tools for analysis.

Generation of tabulations and charts should be at least semi-automatic for standard output - so push-button - with scope for the analyst to dig deeper and produce customised insight - so focusing on what it means, instead of just production of results.


Saul




Reply
#3
"Programming is the 'scripting' of the questionnaire into a survey system it so it will work on the web or by phone for collecting data. However, even pure paper-based questionnaires need to be turned into something that can be used to capture and process the data for analysis."

Hello Saul, and thanks for responding! I'm really trying to understand every part of this. I have a lot of questions, so bear with me.

What does it mean to script a questionnaire? If I'm reading between the lines correctly, it seems as if the MRA, or whoever does that part of it, is taking a format of questions, that has already been answered, and turning them into another format to where the data can be processed.



But, doesn't the paper based questionnaire itself capture the answers from the respondents already? With the question types on it, why cant data be analyzed from the first format? Why does it have to be turned into something else to where the answers can be captured? The answers of the respondent are right there right? I’m pretty sure I'm misunderstanding something.
Reply
#4
Way back in time, the majority of questionnaires were completed on paper. Basically printing documents in bulk that were then completed by pen and pencil (hence the stereotype of researchers with clipboards). However, even here, most professionals would have the paper questionnaires set up on computer and printed out in bulk. This ensured things like columns and codes for punching (when DP also 'punched' or entered data from paper to the computer - originally onto punch cards), and interview instructions and routing commands were set up consistently and correctly.

It's still possible to do pure paper-based questionnaires, but not very common - usually for student learning projects - or a rare postal survey or hotel self-complete 'how did we do?'. At a raw level you create a questionnaire in Word, collect answers, and then enter the data into Excel or a database for counting.

You can even count or do analysis by hand (sometimes still done to check a questionnaire is working at a piloting stage), but electronic analysis is a lot easier. So from the paper questionnaires, the data gets entered into a program that can do the analysis and then tabs - tabulations - are run. The tabs or analysis program can be totally separate from the questionnaire design - eg a stats package, spreadsheet etc. However, most of the time now, design and basic analysis come together in one program.

So while paper is still possible, questionnaires are now mostly electronic - computer-aided surveys such as those online, for phone, or that would be completed via tablet/mobile (computer-aided personal interviewing). It's kind of like creating a specialised type of Google Forms. Although a lot of computer-based surveys are now simple point-and-click (eg SurveyMonkey), scripting - or writing the questionnaire electronically - is still around for more advanced or complex questionnaires, and scripting is still used as a description of moving from paper to electronic even if it is point and click.

Saul




Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)