[I have also posted this on CT.]

This discussion concerns a survey I posted on Crooked Timber (the group blog of which I am a member) and Lifehacker (where I guest-blogged last week).

First, the bullet-point version of this post:

• A one-question survey has very limited utility
• Most respondents have tweaked their default homepage
• Several types of default pages are popular with respondents
• We cannot generalize findings from one blog’s readership to another
• When trying to learn about people’s Web uses, it can be very helpful and interesting to ask them for details

Second, thanks to the 784 readers of CT who took the survey. Read on for more.

The inspiration for this mini survey was two-fold:

1. I was guest-blogging on Lifehacker last week and they have a poll system. I thought it would be fun to poll Lifehacker readers about something. I am interested in people’s Web uses and have been curious for a long time about how common it is for people to change the default setting in their browser and what page comes up when they launch the program.

2. I thought it would be interesting to compare the default browser homepage practices of Lifehacker readers with another group. Since I have access to Crooked Timber readers and CT is also a widely-read blog but with a somewhat different readership, I thought it would be feasible (i.e. I’d get enough responses) and interesting to compare these two groups.

A one-question survey has limited utility

Before trying to take away too much from this exercise, I would like to emphasize that any poll with one question is going to have very limited utility. The poll system at Lifehacker is set up so you can only ask one question. This poses considerable challenges as often we have to rely on multi-part questions to get meaningful answers to what’s of interest. For example, a basic important feature of a survey question is that the answer options be mutually exclusive. It took a bit of tweaking to achieve this here and in the end the results are somewhat hard to interpret with respect to some issues of interest due to the limitations posed by only being able to ask one question.

For example, I wanted to know whether people change the default setting of their browser homepage AND if they do, what they set as their default page. Ideally, I would have liked to collect detailed information about what their default page is if they left it as such, but having all relevant responses made the list of answer choices too long. So the only information I have is whether someone left the default setting, but no information about what that is.

Point being, I am very well aware of the problems posed by a one-question survey of this sort. I would never try to ask a question like this on a survey for a research study. To get at similar information I would break it down into several questions.

Most respondents have tweaked their default homepage

Most respondents – among both Crooked Timber and Lifehacker readers – report changing their default homepage.

Uses default	Crooked Timber (n=784)	Lifehacker (n=1988*)
Yes	9.18	7.65
No	90.82	92.35

Changing the default on your browser is a skill-related question. Although most (all?) browsers make this possible and not necessarily difficult, it is still something you have to 1. realize you can do; 2. decide you want to do; 3. figure out how to do.

We know that blog readers tend to be more frequent users of the Web and tend to be more educated than your general Internet user population so it is not a surprise that most respondents from both blogs report changing their default homepage. MOREOVER, if you have no idea what “browser homepage” means then there is a good chance you would have skipped filling out this questionnaire in the first place.

Several types of default pages are popular with respondents

Click on image for a larger version.

No one type of default setting is used by more than a quarter of respondents. Of course, it is hard to know what comes up as default for those who indicated that they have not changed their defaults. Chances are that those settings lead to a “My Home” (e.g. My Yahoo!, My MSN) type of page, a search engine or an employer/school homepage. So at best, perhaps a third of respondents get a “My Home” or search engine when they first launch their browsers.

A customized “My Home” is considerably more common among Lifehacker respondents than CT respondents. A search engine default home is popular in both groups.

I found this quite curious and already mentioned it in my summary of the Lifehacker poll results over on that blog. As I noted there: “I found the search engine choice interesting. I suspect there are quite a few Firefox users among respondents given the techie background of many Lifehacker readers. A search box is default in Firefox so it would seem redundant to have it on the homepage as well. Various search toolbars exist for other browsers so it’s an option with other programs as well. Then it’s intriguing that people would choose this as their homepage.”

A reader wrote back to explain the choice, which was helpful and interesting. He noted that the advantage of having the search engine come up first is that the cursor is already in the search box so typing in a query does not require an additional mouse movement or keystroke of placing the cursor in the search box. This may seem ridiculous to some (how lazy can we get?), but I can actually sympathize. Still, I don’t start my browsing off with searches often enough to bother calling up a search engine as default and I’m quite happy with using the search box in the toolbar. But that’s just me and understandably other people have different preferences.

Not having anything come up as a default (also known as “blank”) was also popular in both groups although more so among CT respondents.** Many of these people wrote to explain their choice and the most general comment was that they rarely want to go to the same page so why waste time loading up a page? If people are accessing the Web on a relatively slow dial-up connection then this is definitely understandable. For people with high-speed connections, I thought it was curious that this would make such a difference. I have DSL at home – supposedly slower than cable and certainly slower than T1 connections- yet I am rarely inconvenienced by having to wait for a page to load. But to each their own.

CT respondents are more likely to have a news site come up as default, which is not surprising given that CT itself covers a lot of current events so many of its readers are likely the type of people who are interested in avidly following the news, which is made easier by having news sources come up first when they launch their browsers.

About ten percent of respondents in both categories – so representing one of the more popular homepages – have their own page set to come up when they launch the browser. This would explain why people put all sorts of links of interest on their sites. If they use it as a default page then perhaps it functions as a mini-portal. Of course, it may just be that people want to check regularly whether their site is up and running.

A somewhat surprisingly low number of people in both groups reported having Webmail as their first page. Perhaps this reflects few people using this method to access their primary email accounts. (Otherwise, since we know that email use is one of the most common online activities, one would think people would want ready access to the interface.)

We cannot generalize findings from one blog’s readership to another

An important motivation for getting feedback from the readership of two blogs was to see to what extent different blog audiences are similar or, more likely, to what extent they exhibit different types of online practices. We still don’t know that much about blog readers, but a good chunk of what we know – or we think we know – is based on surveys that were posted on some blogs for people to fill out. By showing that responses to one simple question posted on two blogs already exhibit considerable differences, I hope to demonstrate that we have to be very cautious about generalizing from results obtained from one blog’s readership to the characteristics of another blog’s readership.

Of course, this little survey is no different. However, you will notice that I have been very careful in noting the limitations and I do not refer to Crooked Timber or Lifehacker readers when discussing the results, rather, I am careful to restrict my comments about findings to respondents.

When trying to learn about people’s Web uses, it can be very helpful and interesting to ask them for details

This has been a very interesting exercise for me and yet again confirmed for me that instead of trying to guess what people are doing with respect to their online activities, it is much more informative to ask them. In addition to what I already mentioned above, here are some additional interesting tidbits that came out of this exercise.

Especially popular among Lifehacker respondents is SessionSaver for Firefox (or possibly other extensions that allow you to do the same in other browsers). SessionSaver remembers what sites you were on when you closed your browser (even if it quit due to a problem) so you start where you left off.

Some additional interesting starting points included:

• The Breast Cancer Site or the Hunger Site (where you click for donations)
• Either the Wikipedia Main Page or Wikipedia Random Page
• Various picture pages such as the Astronomy Picture of the Day

(There were others, I just mention here ones that several people indicated.)

Some people set up their browser homepages to reflect current events (e.g. Hurricane related information last week) or happenings in their own lives (apartment search page). A few respondents indicated that they create custom files that reside on their own computers and use those as starting points.

Thanks to everyone who filled out the survey and extra thanks to those who took the time to send me individual notes specifying their “Other” choice or mentioning additional observations.

*At the time I post this entry on CT even more responses are showing up on the Lifehacker poll (over 2K at this point). I had to choose a cut-off point so I took the responses close to the time when I posted my summary of the findings there.

**The figures for Blank are not exact. On both blogs I asked people who chose “Other” as their response to follow up with an email to me about specifics. Among CT respondents, 66% followed up, among Lifehacker respondents 40% sent me a note. I then extrapolated from the percentage of these emails referring to “blank” to what it would have been had 100% of those choosing “Other” responded to my query and would have done so in similar proportions with respect to the Blank choice.

This entry was posted on Wednesday, September 7th, 2005 at 9:29 am and is filed under IT/Comm. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.