SURVEY.NET FAQ

What is SURVEY.NET?

SURVEY.NET is a public service network of servers designed to poll users and generate information on demographics, public opinion, trends, etc. SURVEY.NET is based around a customized, ever-evolving script engine developed by Michael Perry (contact).

What makes SURVEY.NET special is that users have the ability to instantly view the latest results of each questionnaire - upon submission of your survey, the system automatically compiles the data and generates a new report, adding your responses to the data pool.

The nucleus of SURVEY.NET is the special engine, which has many advanced features, including the ability to handle user-specified comments (essay-type answers) as well as regular multiple-choice responses. New features are continually being added to the system as more is learned from the process of querying users' opinions and perspectives.

But perhaps more important than anything else, the SURVEY.NET site represents a location users on the web can go and seek demographics, opinions and other information - as well as contribute their own. The underlying engine would be useless if the site itself did not attract a wide variety of internet participants - and this is an aspect that is as critical as anything else - as a result, we strive to keep the site vibrant and alive with opinion and controversy in order to consistently attract users and keep regulars coming back. Each of our surveys is considered a living entity: dynamic; sometimes predictable, sometimes chaotic - just like life. In fact it can be facinating to watch the surveys on various events change in relation to the events unfolding themselves - the OJ Simpson trial survey is a prime example - we captured public opinion before and during each of the side's presentation of their case. The same technique will be used to monitor the events leading up to the 1996 US Presidential election. When scenarios change which affect public opinion, sometimes old surveys will be closed and archived and new ones will be opened - to chronicle not only public opinion in general, but changes in perspectives through time.

Unlike other sites, at SURVEY.NET, the users are the true value - if they're going to submit information, they should have instant access to that information - this is the main reason for making the results automatically available - the data collected is for the enjoyment and education of everyone - and not something devised under the guise of verifying a preset agenda.

If there is an ulterior motive in all of this, it's nowhere near as controversial as our surveys - it is simply to be considered an accurate, reliable source of information on public opinion - the highest honor we can imagine.

Who Runs SURVEY.NET?

SURVEY.NET is run by Mike Perry, an independant software developer and president of Progressive Computer Services (established in 1986). Mike has been involved in developing information systems since the age of 13, creating sites on systems from Control Data's PLATO network, to developing SIGs (Special Interest Groups) on The Source (now CompuServe) and other networks long before the Internet was conceived. Mike has been running a BBS since 1979 and has been awarded "Editor's Choice" in PC Magazine, as well as numerous other honors and endorsements from major publications and writers (including John C. Dvorak) for his programming projects. In the area of marketing, demographics and market research, Mr. Perry has extensive experience in working with advertising agencies, broadcast media and print media - no formal training - just a PhD from the University of Reality in association with services provided to a plethora of very prominent firms.

Ok, this is me (Mike) and obviously I'm also the author of this FAQ - I hate writing about myself in the third person - it's pretty weird, but I'm also not particularly interested in using this forum to blow my own horn (I'd rather be judged by my work exclusively), but I guess these are valid questions and responses to put in a FAQ...

I run SURVEY.NET. I guess you could say it's my baby. It never has been profitable, which IMO is probably the true mark of integrity (ha ha - that's easy to say when you're poor isn't it?). Like everybody else out there, I want to make a mark - I want to do something important, and SURVEY.NET is something I take great pride in.

I'm associated with Inter Commerce Corporation, a local internet service provider - they have been kind enough to indulge my experimentation and attempts to establish this system, and for that I am grateful. If you're looking for an ISP and you're in the area, take a look at what they have to offer. Beyond this friendly suggestion, that's about as much influence as you'll see from me or this site with respect to my hosts; I might have a link to their homepage in places or promos for other local sites, but that's it - anything more would in my opinion be inappropriate considering that we're trying to make this site as objective as possible.

How does SURVEY.NET work? What's the methodology in collecting information?

Obviously, the idea of collecting information online in an automated manner presents some interesting challenges. Traditional polls rely on a 'human factor' and a supposed 'random sample' from which to generate results. It has been argued that traditional polls could be more accurate than the SURVEY.NET system - someone once commented that my system was "biased towards those who are willing to participate." That's funny. I challenge anyone to show me any survey which endeavors to collect accurate information which isn't as biased. The scenario doesn't exist.

Unlike other survey methods, SURVEY.NET is incapable of prejudice. Anyone able to participate is welcome to. There are safeguards to protect against 'spamming' however (see later sections of the document).

The ONLY bias or slant in SURVEY.NET demographics is that our respondents have access to the Internet's World-Wide-Web. If you want to argue whether this group is representative of any sizeable portion of the world society, that's a good topic for debate - but not here. It is my opinion that it's just a matter of time before those who have access to the net will represent a major portion of the world community. Aside from that, we can consider SURVEY.NET's results to be the opinions of those who are on the Internet with WWW access. And as we speak, this segment is getting larger and larger by the minute!

A typical survey in the system presents a small set of questions - usually multiple choice. The results are automatically compiled upon submission of the form. These answers are then added to a "master database" of responses to each question, updated, and then the system re-creates a new report (an html page) reflecting the latest results. The answers to each question are sorted based on the number of respondents, with the most-popular answer being listed first and the least-popular last.

Enhancements are currently underway to implement special variations of surveys which dynamically present new questions based on the results of previous answers - a truly dynamic surveying system... There are also plans to be able to generate results based on specific conditions and responses to specific questions - stay tuned...

How accurate is the SURVEY.NET information? What about multiple users posting answers?

There are numerous techniques implemented to protect against users trying to taint the survey results by completing surveys more than once. For me to elaborate on our protection scheme would defeat its purpose.

In the intial stages of developing the system, we had more than our fair share of people attempt to 'spam' the surveys - it helped us protect against such tactics. In a few cases where users were successful, we were able to identify them in cooperation with their ISP and take appropriate action. There is one person out of a job and one person expelled from a major university as a result of such subversive activities - and even now, they could not repeat their performance as a result of our new protection. Those that were caught that did not have diciplinary action taken against them have had their entire system banned from the survey system - yes we do have an automated blacklist (it's very small though). This isn't to say that it's impossible to defeat our protection against people posting multiple answers, but if you do it, we will catch you and you'll be sorry. In the computer world, ultimately there is no fool-proof security - just like in every other segment of society, people should be left to their own devices to act appropriate and not destroy the work of others. With that being said, at any time we can produce log files identifying the integrity of our surveys.

However, I stand behind the integrity of the SURVEY.NET data and can produce log files which clearly demonstrate that the information comes from unique respondents all over the net.

Are the Surveys anonymous?

YES, YES they are. The system in its present configuration does not keep track of any individual survey answers - never at any time are user responses even written to disk - they are instantly compiled with the main data file. There is no way we can specifically identify what answers any specific user has provided to any specific question - period!

The survey system does log activity however, but it is limited in its scope of what we can identify. The Internet is by its very nature somewhat anonymous - anyone using a browser can fake an e-mail address - so it's very difficult to identify users specifically. We have always had the option of authenticating users and requiring valid information in order to access the site, but we are not implementing this option - for the most part, 99.99999% of our visitors are honorable and honest in answering questions and following the rules - our logs prove this. We don't want to make it a hassel of getting to the site by specifying a username and password - this hasn't become an issue and we hope it never will.

What language is the Survey system written in? What system can run it?

The Survey engine is written in Ansi C and has been tested in a variety of environments - it can probably run on just about any computer that is capable of connecting to the web with a standard Ansi C-compiler available. The source code is about 7,000 lines and growing as we add more features. We have executable code for Novell Unixware, BSD, Sun, and other platforms.

Is the Survey engine available to others?

At the present time there are no plans to make this software publicly available. The main reason for this is that having too many copies of the Survey Engine on the net would dilute the value of our site, and if other people were running more-biased survey systems, it might reflect badly upon SURVEY.NET's reputation as being objective. This is not to say that there are no other versions of SURVEY.NET available online. We're currently working on setting up a European version called "InterPoll" which will be run in the United Kingdom, and of course, being a starving artist, if the money is right, we may consider licensing use of the system. Email me if you are seriously interested in utilizing the technology - we might be able to work something out.

How are new surveys designed? Who decides or creates a new survey?

I pretty-much decide what goes up on the site. I rely heavily on user-input to decide what surveys should be added or how surveys should be modified. I try not to modify a survey after it's been released - at the most I'll correct spelling errors (which I'm slowly becoming famous for (grin)).

The decision to add a new survey is not solely my own - I may have an idea for a topic that I feel may be investigated, but at that point I use the full advantage of the net to seek out other, more-experienced people to help me create a survey. Our current religion survey is a good example. Regardless of my personal religious philosophy, there was no way I or any single individual could come up with an appropriate questionnaire on this topic that would not alienate particular groups - I contact the an agency, the Ontario Centre for Religious Tolerance and asked for their guidance in helping me develop an appropriate survey. Quite often, the intial survey is run for a short period of time; I get feedback, revise the survey and start a new one with enhanced questions.

If I have an idea for a survey, what should I do?

Send me e-mail. Most of what's online is the result of user feedback so please feel free to offer your suggestions for new surveys or enhancements to existing surveys.

The more organized your response is, the more-likely it will become a reality. If you present a well-thought-out list of questions & answers, there may be a good chance it could be the basis for a future survey. BUT, please be advised that we primarily publish surveys of a general nature - if you're looking for something specific, there's not a great chance that we will run a survey.

What if I wanted you to run a specific survey? Will you do that for organizations?

Yes. It depends upon the nature of the survey however, as to whether or not it would be included in the main SURVEY.NET system. There is a standard which the main system adheres to. Being a starving artist, I'll gladly produce surveys for commercial entities, especially if I can help subsidize my activities, but if it is something in conflict with our principals of objectivity, I may ask that the survey be isolated and promoted independetly. Along this line, I'll also entertain offers to license the Survey Engine for use on other systems by other organizations - this use will be limited in most cases - either for private, non-competitive use, or in accordance with the Survey Net standard of being as un-biased as possible.

Recently, I was contacted by the folks at MegaMall who wanted to accumulate some demographics & opinions on online shopping - obviously this is something of great value so we worked together to create an appropriate survey. Everyone is now able to exploit the information generated. I'm particularly interested in major sites sponsoring surveys - especially if they direct traffic to SURVEY.NET - the more people the merrier!

If my company wanted to license your software or have something customized done, is it possible?

Yes. You'll have to contact me, but I'm open to a variety of ideas on this. Contact me here. I develop net-based client-server applications in my business, and if you like SURVEY.NET, then you have an idea of the quality of my work. I also have a number of other advanced systems which have been created after SURVEY.NET, including "The Engine", an automated shopping mall system, and WebSet - a program that can be used to create web pages using "templates" from a browser. WebSet is currently being implemented in SURVEY.NET to allow authorized columnists to automatically update sections of the SURVEY.NET site - in this manner the authorized writers can update their own pages without any intervention.

Can the survey results be republished? Are there any restrictions?

All information generated by SURVEY.NET is republishable in any form. We ask that you give us credit in your reference in return - that's all. If possible, list our WWW URL of "http://www.survey.net/index.html, or credit "source: Mike Perry/SURVEY.NET". Other than that, feel free to re-print the information. Please endeavor to reprint the information in it's entirety however - we do not want the data to be extrapolated and used out of context. It's ok to extrapolate responses to individual questions, but do not endeavor to mis-represent the data accumulated. We reserve the right to restict any organization's ability to republish data from our site if it is distorted in its presentation. We ask that any republished information be sent to use (primarily for my own amusement) to: PCS, PO Box 7638, Metairie, LA 70010

More information can be found on the License Page.

I've noticed some disparity between the number of respondents and total survey answers to specific questions, what gives?

This disparity seems to be the result of certain types of browsers not completely sending all forms data to the script - it's beyond our capability to deal with. We try to keep the number of survey questions limited so that those who are running 'buggy' browsers don't make the numbers not properly add-up. This doesn't affect the accuracy of survey data - however there may be some cases where some users' answers to specific questions are simply not recorded. It's very difficult to track down the cause of this problem - it is noticable when you get into the area of having a thousand or more surveys completed - so it's not very common. Different browsers handle forms data in different manners - some browsers will NOT send a survey answer if the user didn't specifically pick the response (and instead left the default answer as it was). If you're analyzing survey data, assume any disparity in the number of answers to a specific question to be defined as "No answer."

In some RARE cases, the disparity may be due to the survey being modified after it was released and the results not cleared. If I run across a survey where I feel an additional question should be added, I may add the question/answers to the survey - I've only done this once and will not do it often. And in such a case, the rule above applies. If 75 people answered a survey and I decide to add a new question, you may notice that one question has 75 less responses than all the others - just consider those 75 to be null or "no answer" since the question wasn't there before. This will very rarely occur. I'm more apt to reset an entire survey if I make any major changes.

The END (for now)