LimeSurvey 2.0 user types

= overarching personas=

(by brian o'neill based on talks with IQSS staff, meetings, and talk with a marketing manager Jan 2008)
 * 1) *Marketing Manager* - Mary Marketing uses LimeSurvey but probably didnt do the installation and has little to no technical skills beyond basic office/DB management stuff for their customers. She is most interested in being able to gather reports on multiple-choice question surveys, and peruse the open/prose commentaries from her customers. She is not interested much in survey methodologies and the results of her work are usually in the form of delivering to her boss the results of her efforts in an executive office. She has a customer relationship management (CRM) tool and a database of customers she would like to link to the tool. She runs certain surveys regularly but also creates new surveys occasionally to test new product ideas. She may also use the tool to create INTERNAL employee surveys alongside the HR department. Frequencies: customer sat surveys generally are 1x yr only. 4x/yr for high-value, willing customers. (People have a low threshold for taking surveys). The customer sat survey will change slightly overtime but will often be the same questions year to year ; randomization and basic anti-biasing features would help this user. They would care a lot about baselines and trending over time (year to year comparisons of sat surveys). They are more accustomed to filtering questions after the fact (E.g. send everyone in a group the same questions and filter out the ones that dont make sense; vs trying to learn/understand how to hide questions based on previous historical data). These might be the people on LS currently begging for more question types. Time is money to this person; too much hassle and they will pay somebody  (budget allotting). in the past, all surveys were budgeted out to external companies but at some point, time becomes money for them. The BIGGEST time suck is in the REPORTING aspect of the surveys.
 * 2) *Anonymous survey taker:* this person knows nothing about the workings of lime survey at all and simply completes a survey.  These are people like product consumers or academic/R&D; engineering staff reporting on compliance. Their scope of knowledge is bookended by a welcome screen for a survey, and a thank-you page at the end with email being the method of informing them of a survey. Commonly, this person does not have a real name in the system and is identified by an email address. This is the person answering a best buy sat survey after having bought a TV on their website. <?>
 * 3) *"Aware" survey taker(employee of a company, academic/R&D; engineer, etc)* - this person knows they are using the LS sofware package; they may know they are part of a team completing required forms/surveys on a routine basis. They may have access to, and care, about the reporting on the survey results unlike the anonymous taker. Their account has a real name most likely associated with it.
 * 4) *Academic survey researcher / scientist* - this would be analagous to the IQSS department at harvard. These are users with advanced domain knowledge. They care a lot about the exporting formats for the data, which they want to be extensive. they are not interested in elementary reporting aspects of LimeSurvey and want to crunch the data themselves. They may have programming experience and the ability to write their own plugins such as anti-biasing rules to questions. <?> They need good panel management. They probably know about datavers and may want to decouple certain aspects of LS's out of the box settings in favor of their own external tools. They probably care a lot more of the effects of testing bias and other external stimuli than they do the actual results of the data - and they may be the final recipient of the data if the entire purpose of the study was to test things beyond what the user may think they are being tested on.<?>
 * 5) *Oversight/Executive/Manager* - this person cares about the results of the survey; they may have limited interaction with the LS platform in any way; they may want to log in and view the latest info on how soon the survey will end, the latest results, and perhaps building a report using the embedded LS tools. Most of the academic/R&D; oversight people probably fall into this category.
 * 6) *Survey Researcher/Developer, Research Department of Large Corp* - this person is responsible for running tests and surveys for the marketing (and other groups) within a company. They have more technical expertise than the marketing manager does and have formal knowledge of the domain. They may have programming experience <?> This person probably doesnt deal with panel management and administering the survey and is more interested in developing and reporting on the survey.
 * 7) *LimeSurvey Software Engineering Contributor:* this person has extensive coding knowledge and may be a super version of the persona above. They have the ability to add on to the codebase of LS and know how to make plugins etc. They care about not being left in the dust and want to have input into the future visions of the product.

Personas a la Caroll
Persona Work In Progress

=Survey Analysis of 'Lime Survey Use and Future Directions v2 (ID 42159)'=

Results from 9500+ responses to the feedback-questionnaire of limesurvey are analyzed here.

In case you want to know more about the survey: http://survey.limesurvey.org/index.php?sid=42159

Language-Use


A lot of participants preferred an English questionnaire to a German one.

LimeSurvey Usage Sectors
Sector of Usage: Commercial 37.4%, Public 38.3%, Private 33.7"



LimeSurvey users are using this tool slightly more often for commercial and public purposes.

There is no preference between commercial or public usage.

What do people use LimeSurvey for?
Purpose of Usage:
 * Data Entry 33.9%,
 * Market Research 40%,
 * Option polling 41%,
 * Organising Purposes 15.8%,
 * from a total of 773 responses



Most People use LimeSurvey for Market Research or Option Polling.

However, note the low number of Responses (773 out of 9500+), maybe the question is ill-defined?

Most responses from "Other" relate to an academic project as i.e. "for my master's thesis".

All interesting responses:
 * Fehlberseitigung in LimeSurvey(Error correction in LimeSurvey)
 * Umfragen (Questionnaires)
 * psychologische Testkonstruktion (Psychological Testconstruction)
 * Part of a commercial product
 * client assessment questionnaire
 * Research
 * infrastructure questions
 * employee research
 * Methodology teaching
 * Training Evaluations
 * Transport surveys
 * Assessment
 * Meinungen der User einzuholen; Website/Angebot verbessern (to a improve Website)
 * science
 * Graduate Surveys
 * Large group data collection, survey
 * registration form
 * Event registration
 * Customer Qualification

Rating LimeSurvey Features
From a total of 888 Responses:



Interpretation:
 * The ease of use of the Admin-Interface is rated rather bad (41% less than ok).
 * The ease of use for the participants is rated slightly better with much less "terrible" responses (10% vs. 20% above).
 * Templates and Statistics are rated most useful
 * Export is rated least useful, but differences are quite small between usefullness of Export, Conditions, Statistics and Templates

Installation, who did it?
From total 747 responses:
 * Myself 67.3%,
 * Administrator 36%,
 * Friend with IT knowledge 10.3%



Which Additional Feature do you need:


Other Features:
 * more question types (3 times)
 * better confition branch (testing for number, OR, XOR ..;) (3 times)
 * Pie charts, graphs, and charts (3 times)
 * simplification of reporting results (2 times)
 * Add graphics to questions (2 times)
 * better, XHTML templates (2 times)
 * Dual Scale Matrix
 * speed?
 * API (not sure if it exists, haven't searched for it tho)
 * Plug-ins for result evaluations
 * export to MS Access as SQL
 * pre-populated answers
 * API to control LimeS from external application (for instance managing/querying tokens,) (3 times)
 * make authentication more modular (ldap, mysql, ntauth authent) for both survey admins and participants, ...
 * use answers already given as text strings in subsequent questions
 * A nicer and more intuitive administration interface.
 * Help files...
 * Multiple grid answers choice
 * PDF export, Export in OpenOffice Format                                                                                                                                                                                                               * A data entry mode that allows an import of certain fields, and then an easy way to access individual records to finish each record.
 * Easier documentation for Survey Developers
 * Interface for custom surveys (e.g. external pages that feed results into survey)
 * Different ending for each branch
 * inventory of questions for using the again and so have timelines through surveys                                                                                                                                           * Automatic score calculation for tests
 * Ability to Charge Clients
 * copy survey without export-reimport
 * Export to SPSS
 * Attach and upload a file
 * Super Admin
 * User permissions
 * multiple registrations per participant
 * require access code to complete survey
 * postgresql
 * robot access control
 * more free custom attributes for invitations
 * multiple options array question format
 * Conjoint Analysis, Inline Multimedia, Looping

Priority ranking of improvements:


summary:
 * users want improved conditioning
 * having more templates and admin interface improvements are perceived as equally important
 * we have enough localization ;)

Other comments:

 * very low signal to noise ratio, still working on this

material:

 * R-script to analyse the data: [[Media:urvey_analysis.R_des|R-Script]]
 * Datafile from the survey in csv format: ...