 |
|
NONE: Re: ONLINE-ADS>> Surveys, cache busting, and cookies
Re: ONLINE-ADS>> Surveys, cache busting, and cookies
Steve Runfeldt (Steve_at_CustomerSat.com)
Thu, 19 Feb 1998 12:49:03 -0800
Glenn Fleishman wrote:
> This misses the point. Any well-conducted phone survey attempts to gather
> some number of people with certain characteristics, whether race,
> ethnicity, gender, age, time spent online, membership in certain
> organizations, etc. If a telephone surveyer is turned down (which probably
> happens most of the time), they call another person - they don't reduce the
> size of the sample by one.
>
> The point of methodology and building some kind of relationship between the
> people who participate in a survey and the conclusions that are drawn is to
> a) allow others to analyze the results, b) have apples-to-apples
> comparisons, and c) provide an estimate of accuracy or error rate.
>
These are very good points. We often use screener questions and can employ
quotas
on a Web survey. The main reason for doing apriori selection on a phone
survey is
cost. You don't want to spend the phone interviewers time surveying
unnnecessary
people. With a Web survey, there is virtually no added cost (or very
little) for
additional respondents, so the selection can be made post hoc, either by
weighting
or post-hoc quotas. We can and often do also use screener pages which can
be quota
based. Again, I am not talking about open Web page surveys like the GVU
project,
but more scientifically controlled approaches.
The bottom line is that we are getting between 25 and 50% response rates on Web
surveys. With percentages like that, the self selection objection becomes moot.
The effect of self selection is much greater on the phone. With a phone
survey, if
you catch the person at the wrong time you lose them. With email or the
Web, the
respondent chooses their own time to respond. If the sample is of people
with whom
you already have a relationship, the rates are very high. If the market is the
Internet, it just makes sense to survey on the Internet.
> I'm not saying that polls or surveys are perfect when they're conducted by
> phone. But I am saying that statistical science underlies good market
> research and polling.
>
Amen.
> >This approach, too, provides
> >much better sampling than a simple open invitation to a
> >survey.
>
> Too easy and more likely for people to lie on the Web, although I agree an
> invitation to a percentage of visitors (based on behavior, preferably) is
> superior to a general call to fill out a survey with no limitations.
>
Our experience suggests that there is actually more of a tendency to lie to
a live
interviewer than to a Web page. It may depend on the survey and how it is
presented. If, for example, people are required to fill out a survey in
order to
gain access to a Web page, it is likely that many will lie. The GVU survey
results
support this. If you are surveying an established customer base and have
given the
respondents the option to not respond at all, the incentive for giving false
information is not there.
On the other hand, respondents talking to a live interviewer may be
predisposed to
hold back on negative views. No one wants to give poor ratings. People
tend to be
more candid on the Internet. On customer satisfaction ratings we tend to see
lower mean scores and greater variance on Internet questionnaires than from
phone
surveys. This suggests greater honesty rather than less. The verbatims we
receive
are incredibly rich. People who wouldn't say much on the phone will write pages
and pages in an email or comment box. Just look at this listserve for an
example
of that.
It is important that surveys be designed with the appropriate scientific and
statistical safeguards. It is also important that the methodology match the
market. If the customer base is on the Internet, then Internet surveying makes
much more sense than more traditional methods.
Regards,
Steve Runfeldt
VP Research
CustomerSat.com
----------------------------------------------------------------------
This week's Online Advertising Discussion List sponsors:
HotBot and ICONOCAST
HotBot, rated the #1 search engine by CNET, has keywords available now
in these categories: computers, software, shareware, Internet, games,
music, and more. Contact Rick Boyce at 415.276.8440 or rick_at_wired.com.
---
"ICONOCAST is without question, the coolest, shortest,
best written Internet thing I read every week."
Seth Godin, CEO, Yoyodyne http://www.iconocast.com
-----------------------------------------------------------------------
Online Advertising Discussion List To Unsubscribe send UNSUBSCRIBE
http://www.o-a.com/ to online-ads-request_at_o-a.com
HOW TO JOIN THE ONLINE ADVERTISING DISCUSSION LIST
|
With an archive of more than 14,000 postings, since 1996 the
Online Advertising Discussion List has been the Internet's leading forum focused on professional discussion
of online advertising and online media buying and selling strategies, results, studies, tools, and media
coverage. If you wish to join the discussion list, please use this link to sign up on the home page of the Online Advertising Discussion List. |
|
|
Online Advertising Industry Leaders:
Clicksor
List and Found
AdJungle
The Laredo Group
Add your company...
|
