Re: [Swprograms] Why Researchers Should Be Gelded

[Richard Cuff <rdcuff@xxxxxxxxx>]

Well put, Mike.

The second study should be viewed with a more jaundiced eye because it
is a projection, and a 9-year-out projection at that!  While
projections that far into the future might be useful for planning
infrastructure developments, consumer tastes are notoriously fickle
and not projectionable beyond basic trends.  Whether or not one
person, let alone a bazillion people, are interested in satellite
radio in 2014 will largely depend on the options available to people
in 2014 and what type of content exists.  Throw in the potential
influence of broadcast regulators such as the FCC and CRTC, both of
which are motivated by matters other than altruism, and you have
further muddied the waters.

Considering that satellite radio was probably rather hazily defined
four years ago, I wouldn't put too much stock in that first study,
either.  I'm more interested in the market interest -- more
imporantly, the intention to purchase -- from two years ago vs. today.
 The satellite radio product has been reasonably static in that time.

The discipline of statistics is a very useful analytical tool. 
However, improperly applied, and improperly interpreted, a lot of
hooey can result.  One must always consider the organization who
originated the report, their agendas, and the agendas of those who
commissioned the study.

Market researchingly yours,

Richard Cuff / Allentown, PA

On 7/22/05, Mike Barraclough <softbulletin1@xxxxxxxxx> wrote:
> I don't see how Twain's comment adds to this debate.
> It's based on mathematical ignorance.
> 
> The surveys are using two different statistical
> methods of looking at the issue, projections would
> always be less reliable. With satellite radio you only
> have a few years figures with which to construct your
> regression line or curve and, as the researchers
> admit, you can't build into that the effect of other
> technologies.

_______________________________________________
Swprograms mailing list
Swprograms@xxxxxxxxxxxxxxxx
http://dallas.hard-core-dx.com/mailman/listinfo/swprograms

To unsubscribe:  Send an E-mail to  swprograms-request@xxxxxxxxxxxxxxxx?subject=unsubscribe, or visit the URL shown above.