Seeking Alpha
View as an RSS Feed

J. M. Manness  

View J. M. Manness' Comments BY TICKER:
Latest  |  Highest rated
  • Apple Watch Edition: Price - Simple Logic [View article]
    Tales:
    "Given the teardown I reiterate my prediction that there will be no Watch upgrades beyond batteries..."

    Looking over the review carefully I am not sure that what you say holds true. While "pulling out this mess is a destructive procedure..." it hold be noted that if you want to replace the S1, then destroying the old one is not an issue.

    Additionally, if the soldered connections were initially made, then they can be either undone or cleaned for reuse. Finally, if the whole internals are to be replaced, then whatever these are soldered to will be replaced as well, so any mess is irrelevant. Obviously, there is no motivation to repair the Sport model where replacement parts and repair would be more expensive than replacing the whole thing.

    But with a case with $900 worth of gold, what is an hour our two labor to do the job.

    So I would argue that the jury is still out on this one.

    Regards.
    Apr 24, 2015. 02:16 PM | 1 Like Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    @lolsz

    Finally - I would like to remind you that Slice is NOT trying to say anything about the USA population as a whole, nor predict any future sales, they are merely extrapolating a reasonable figure for the number sold and average price points.
    Apr 15, 2015. 10:18 AM | Likes Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    If you work in market research - be kind enough to detail to us how you would select truly representative sample of buyers of Apple Watches.
    Apr 15, 2015. 10:13 AM | 2 Likes Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    By your definition we would never ever have a truly representative sample in the field of human activity since almost no sample no sample is ever completely random. (This would not include selecting from a set of the whole domain - e.g. ALL sales receipts for a given store, all IRS filings - but this is not possible if one does not have access to the whole population. Even if you do have all sales receipts from a given store you will only be able to do valid statistics for those sales - you can say nothing about those who did not buy.

    People who do political surveys work with much smaller samples sizes by deliberately electing from various population groups.

    I have never said that what you raise is not an issue - only that one needs to do 2 things.

    1 - keep in mind that the representativeness of the sample may be biased and

    2- Adjust for it if possible.

    The reader should also evaluate the source of the data and evaluate to what extant the non-representativenss of the sample may bias the outcome.

    ANY human survey is biased to some degree because you only get people who willing to participate.
    Apr 15, 2015. 10:12 AM | 1 Like Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    Iseeyou...

    Apple is ALWAYS facing manufacturing issues. They must start up a complex production system that uses numerous technologies that have not been done before or not in this particular way.

    All this latest tech has issues to be worked out and effects timing all along the production process. Watch the intro videos to see just a small part of it. It is a real wonder that any of it gets done in time for launch at all.

    IMHO

    regards!
    Apr 14, 2015. 08:11 PM | 2 Likes Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    The question is....

    Did China buy up ALL the Edition watches and that is why they did not show up on Slice's USA only data?

    It is a plausible scenario.
    Apr 14, 2015. 01:56 PM | Likes Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    Doing occasional amateur video (and a lot of photography) I can relate - but not justify the cost for me.

    Thanks for the comment
    Apr 14, 2015. 01:52 PM | Likes Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    Thanks Jaimee - I appreciate your continued dialog here.
    Apr 14, 2015. 01:02 PM | Likes Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    @dollars

    Gary J was just being "cute" I did get a chuckle out of it.

    I am sure that you can cancel any time before shipment with no penalty. Once shipped, I am not sure what they would charge for a restocking fee.
    Apr 14, 2015. 12:28 PM | Likes Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    @xyz

    no critical comments have been removed (to my knowledge). I certainly have not request any.

    I will request removal of rude or insulting remarks. Your "kook-aid" comment verges on that.

    Regards
    Apr 14, 2015. 11:46 AM | 2 Likes Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    As noted above - a sample does NOT need to be random to be representative.

    In fact, Slice's sample indeed is random in that there is no pressure put on an individual to join or not, and it is open to all.

    In any case, The question is not "is it representative of USA population?" but rather "Is it representative of USA electronics population?" AND can we move from this sample to a REASONABLE ESTIMATE of total sales.

    You write:
    "Reader beware. methodologies used to give advice should be checked out thoroughly before you make any buying or selling decision."

    I agree whole heartedly! I have consistently advocated for this, such as my post on Misreading data:
    http://seekingalpha.co...

    Thanks for the comments.

    Regards - JMM
    Apr 14, 2015. 11:28 AM | 1 Like Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    lolsz

    You are of course correct that it is not truly random. But RANDOM is not what is necessary -REPRESENTATIVE is. So the question should be"
    "To what extent is the sample representative?"

    Obviously we can never know for sure. Given the size however, and the fact that buyers of preorders are by definition similar in that they all purchase online and live in USA, my guess it that it is reasonably representative, and that Slice makes great efforts to account in their model for discrepancies when they get feedback from vendors. (Although I imagine Apple is not doing that.)

    Regards - JMM

    (Caveat - I am not a statistician but did study some stat at University level)
    Apr 14, 2015. 11:16 AM | 2 Likes Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    @JP
    "any future updates planned?"

    To my article? What would you like to see?
    Apr 14, 2015. 11:08 AM | 1 Like Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    "...like trained seals."

    ARRR ARRR ARRR ARRR

    :-)
    Apr 14, 2015. 11:05 AM | 1 Like Like |Link to Comment
  • Apple Watch: Understanding Early Sales Data [View article]
    giofls

    Good question.

    First - it is not MY data - but from Slice.com. (You may want to check out their service just to see how it works.)

    Their "domain sample" is clearly subscribers to their service. As you note, this is a somewhat selective sample. However, since the whole field is one of technology, and all ordering was indeed online, then I think this is less skewed than it might otherwise be.

    Finally - note my quote "Based on our statistical modeling and validation (including partner and client feedback), we are confident our numbers are representative. "

    So they have their proprietary formula for moving from their data to a projection.

    Each individual will need to decide how he/she will rate that projection. My guess is that it will end up being reasonably accurate.

    Regards - JMM
    Apr 14, 2015. 11:03 AM | 2 Likes Like |Link to Comment
COMMENTS STATS
2,236 Comments
3,393 Likes