fluid-work Digest, Vol 22, Issue 15

Herb Wideman herb at yorku.ca
Wed Dec 10 17:30:59 UTC 2008

Hi Blake,

I don't know if you have received any feedback from Ron on your question as he is in Asia, but based on our experience using surveys I would second the point raised earlier about the need for a matrix-type question, which is commonly used for a sequence of questions that all have the same answer options - it creates a very neat and space-saving tabular layout format for the survey taker. This format is heavily used in surveys so I would give it high priority. Yes/no dichotomous (boolean) questions are rarely used so I would lower their priority for development.  The rating question type you have at the end of your list can be subsumed as a variant of the single-option question type by allowing the creator the option of setting numeric values for each of the selectable test answers they create in the answer field. These numeric values are not seen by the respondent but are placed in the answer file in lieu of the text labels for the values. This option is very important to us as it allows easy export of quantified data for statistical analysis without the need for additional conversions (of the type very easy=1, easy=2 etc). Again, see the York U survey entry tool to see how this can be implemented.


 Hello All,
>> Blake here from Toronto/VULab WEB, and if you haven't herd - we are  
>> making the switch to a proper php framework (EXCITING!). We are  
>> looking at CodeIgniter.com and CakePHP.org, but that isn't why I'm  
>> calling on you this morning.
>> Due to the dramatic (positive) change that a framework will bring  
>> to VULab, we will be re-engineering the survey design tool/process  
>> (nearly) from the ground up!
>> I want to start by isolating the types of questions we want to  
>> offer as well as isolating the importance of each. Once we have a  
>> list of question types we can agree on, we can develop them into  
>> VULab working down the list from most important to optional  
>> questions. You should consider what type of information you'd want  
>> to gather if you were conducting a usability test.
>> VULab currently uses the phpESP survey design tool so I'll outline  
>> what is included there and then I'll give me opinion on types/ 
>> priorities.
>> One thing that should be noted is that a question "type" can span  
>> different "implementations". For example, a Boolean type of  
>> question could be implemented as a True or False, or a Yes or No,  
>> or even an Apples or Oranges. I'll distinguish between the types of  
>> questions and the implementations below.
>> phpESP Current Question Types:
>> Boolean Question Type
>> This is a question type whose answers could resolve into a 1 or 0.
>> Implementations:
>> Yes/No
>> Text Input Question Type
>> This is a question type whose answers are stored as text.
>> Implementations:
>> Text Box
>> Essay Box
>> Numeric
>> Single Option Question Type
>> This is a question type whose answer is chosen from a list and is  
>> specific to one option.
>> Implementations:
>> Radio Buttons
>> Dropdown Box
>> Multiple Select Question Type
>> This is a question type whose answer can be 1 or more options from  
>> a given list.
>> Implementations:
>> Check Boxes
>> Date Question Type
>> This is a question type whose answer is a specific date.
>> Implementations:
>> Date
>> Rating Question Type
>> This is a question type whose answer is a number within a selected  
>> scale.
>> Implementations:
>> Rank.
>> As I see it, we can consolidate a few of these and prioritize them  
>> as follows for VULab
>> Note: Single/Multiple types have been combined.

More information about the fluid-work mailing list