Ontolly wrote:The angriest I've seen our lass was the morning after the night I'd been a little bit of a bender in Cardiff.
Whatever floats ya boat fella
Ontolly wrote:The angriest I've seen our lass was the morning after the night I'd been a little bit of a bender in Cardiff.
MightyWhite wrote:Ontolly wrote:The angriest I've seen our lass was the morning after the night I'd been a little bit of a bender in Cardiff.
Whatever floats ya boat fella
Ontolly wrote:MightyWhite wrote:Ontolly wrote:The angriest I've seen our lass was the morning after the night I'd been a little bit of a bender in Cardiff.
Whatever floats ya boat fella
I see what you did there. Very good.
Ontolly wrote:Being asked to look at online surveys, Being shown video evidence of your husband pissing through the letterbox of Costa coffee - I guess it's all relative. There's just no pleasing some women.
right so the main methodological issues that make this survey design utterly absurd and the results near impossible to interpret in any sort of meaningful or reliable way are relatively apparent but basically this dude's survey suffers from two major critical flaws - a near complete lack of any sort of clear or concise conceptualization of his dependent variables and a measurement approach so sloppy and disfigured he'd probably have better luck just making up answers than he would distributing this to real people. the survey fails the tests of reliability (the ability of the survey to replicate the same results with repeated tests) and validity (the ability of the survey to measure accurately the variables it purports to measure) for the following reasons -
- double-barrelled questions which ask more than one question or make more than one statement in the same item, a fatal flaw which totally and utterly discounts the possibility that respondents may not share the same opinion about all statements and effectively sends the validity of your data straight into the garbage
- leading questions which destroy the validity of the researcher's data because in asking questions that hint at the answer the researcher wants to hear he's essentially constructing and fulfilling a self-fulfilling prophecy that contributes little to nothing new in the way of reliable, unbiased knowledge
- presumption that all respondents responding to the survey do not have anything but utterly positive views about their team - i hope i don't have to explain why removing entirely any and all options to express negative views on a survey biases and mangles that data to the point where you sincerely might as well just be making up whatever shit you want to hear
if this dude really isn't sure how to conceptualize his variables because he's working from the ground up and doesn't know what kind of framework is going to emerge from his data then the best thing he can do is just isolate and treat each statement as a separate dependent variable, measured on a likert scale for manageable quantitative analysis. once he's got his dataset, he can run a principle components or factor analysis to eliminate the redundant intercorrelations and overlaps in his responses and condense the whole questionnaire into a much smaller set of variables that he can then name and analyze in whichever way he sees fit. of course, the big question is that of how the fuck any university allowed this dude to do a serious dissertation while lacking the fundamentals of how to both design and meaningfully analyze an efficient and methodologically sound questionnaire (i hate to sound patronizing here but from a researcher perspective this really is basic stuff, maybe intermediate at best), which remains utterly beyond me
Do you even like me. what the fuck is this
tommydski wrote:Later on, she sent this too -
right so the main methodological issues that make this survey design utterly absurd and the results near impossible to interpret in any sort of meaningful or reliable way are relatively apparent but basically this dude's survey suffers from two major critical flaws - a near complete lack of any sort of clear or concise conceptualization of his dependent variables and a measurement approach so sloppy and disfigured he'd probably have better luck just making up answers than he would distributing this to real people. the survey fails the tests of reliability (the ability of the survey to replicate the same results with repeated tests) and validity (the ability of the survey to measure accurately the variables it purports to measure) for the following reasons -
- double-barrelled questions which ask more than one question or make more than one statement in the same item, a fatal flaw which totally and utterly discounts the possibility that respondents may not share the same opinion about all statements and effectively sends the validity of your data straight into the garbage
- leading questions which destroy the validity of the researcher's data because in asking questions that hint at the answer the researcher wants to hear he's essentially constructing and fulfilling a self-fulfilling prophecy that contributes little to nothing new in the way of reliable, unbiased knowledge
- presumption that all respondents responding to the survey do not have anything but utterly positive views about their team - i hope i don't have to explain why removing entirely any and all options to express negative views on a survey biases and mangles that data to the point where you sincerely might as well just be making up whatever shit you want to hear
if this dude really isn't sure how to conceptualize his variables because he's working from the ground up and doesn't know what kind of framework is going to emerge from his data then the best thing he can do is just isolate and treat each statement as a separate dependent variable, measured on a likert scale for manageable quantitative analysis. once he's got his dataset, he can run a principle components or factor analysis to eliminate the redundant intercorrelations and overlaps in his responses and condense the whole questionnaire into a much smaller set of variables that he can then name and analyze in whichever way he sees fit. of course, the big question is that of how the fuck any university allowed this dude to do a serious dissertation while lacking the fundamentals of how to both design and meaningfully analyze an efficient and methodologically sound questionnaire (i hate to sound patronizing here but from a researcher perspective this really is basic stuff, maybe intermediate at best), which remains utterly beyond me
MightyWhite wrote:tommydski wrote:Later on, she sent this too -
right so the main methodological issues that make this survey design utterly absurd and the results near impossible to interpret in any sort of meaningful or reliable way are relatively apparent but basically this dude's survey suffers from two major critical flaws - a near complete lack of any sort of clear or concise conceptualization of his dependent variables and a measurement approach so sloppy and disfigured he'd probably have better luck just making up answers than he would distributing this to real people. the survey fails the tests of reliability (the ability of the survey to replicate the same results with repeated tests) and validity (the ability of the survey to measure accurately the variables it purports to measure) for the following reasons -
- double-barrelled questions which ask more than one question or make more than one statement in the same item, a fatal flaw which totally and utterly discounts the possibility that respondents may not share the same opinion about all statements and effectively sends the validity of your data straight into the garbage
- leading questions which destroy the validity of the researcher's data because in asking questions that hint at the answer the researcher wants to hear he's essentially constructing and fulfilling a self-fulfilling prophecy that contributes little to nothing new in the way of reliable, unbiased knowledge
- presumption that all respondents responding to the survey do not have anything but utterly positive views about their team - i hope i don't have to explain why removing entirely any and all options to express negative views on a survey biases and mangles that data to the point where you sincerely might as well just be making up whatever shit you want to hear
if this dude really isn't sure how to conceptualize his variables because he's working from the ground up and doesn't know what kind of framework is going to emerge from his data then the best thing he can do is just isolate and treat each statement as a separate dependent variable, measured on a likert scale for manageable quantitative analysis. once he's got his dataset, he can run a principle components or factor analysis to eliminate the redundant intercorrelations and overlaps in his responses and condense the whole questionnaire into a much smaller set of variables that he can then name and analyze in whichever way he sees fit. of course, the big question is that of how the fuck any university allowed this dude to do a serious dissertation while lacking the fundamentals of how to both design and meaningfully analyze an efficient and methodologically sound questionnaire (i hate to sound patronizing here but from a researcher perspective this really is basic stuff, maybe intermediate at best), which remains utterly beyond me
To which your response was "Ok duck, put kettle on. There's a good girl".
I hope.
tommydski wrote:...Later on, she sent this too -
right so the main methodological issues that make this survey design utterly absurd and the results near impossible to interpret in any sort of meaningful or reliable way are relatively apparent but basically this dude's survey suffers from two major critical flaws - a near complete lack of any sort of clear or concise conceptualization of his dependent variables and a measurement approach so sloppy and disfigured he'd probably have better luck just making up answers than he would distributing this to real people. the survey fails the tests of reliability (the ability of the survey to replicate the same results with repeated tests) and validity (the ability of the survey to measure accurately the variables it purports to measure) for the following reasons -
- double-barrelled questions which ask more than one question or make more than one statement in the same item, a fatal flaw which totally and utterly discounts the possibility that respondents may not share the same opinion about all statements and effectively sends the validity of your data straight into the garbage
- leading questions which destroy the validity of the researcher's data because in asking questions that hint at the answer the researcher wants to hear he's essentially constructing and fulfilling a self-fulfilling prophecy that contributes little to nothing new in the way of reliable, unbiased knowledge
- presumption that all respondents responding to the survey do not have anything but utterly positive views about their team - i hope i don't have to explain why removing entirely any and all options to express negative views on a survey biases and mangles that data to the point where you sincerely might as well just be making up whatever shit you want to hear
if this dude really isn't sure how to conceptualize his variables because he's working from the ground up and doesn't know what kind of framework is going to emerge from his data then the best thing he can do is just isolate and treat each statement as a separate dependent variable, measured on a likert scale for manageable quantitative analysis. once he's got his dataset, he can run a principle components or factor analysis to eliminate the redundant intercorrelations and overlaps in his responses and condense the whole questionnaire into a much smaller set of variables that he can then name and analyze in whichever way he sees fit. of course, the big question is that of how the fuck any university allowed this dude to do a serious dissertation while lacking the fundamentals of how to both design and meaningfully analyze an efficient and methodologically sound questionnaire (i hate to sound patronizing here but from a researcher perspective this really is basic stuff, maybe intermediate at best), which remains utterly beyond me
the flying pig wrote:
i warrant that mrs dski would be sneering on the other side of her face if she had even half an inkling of what the big man is packing in the trouser department
if this dude really isn't sure how to conceptualize his variables because he's working from the ground up and doesn't know what kind of framework is going to emerge from his data then the best thing he can do is just isolate and treat each statement as a separate dependent variable, measured on a likert scale for manageable quantitative analysis. once he's got his dataset, he can run a principle components or factor analysis to eliminate the redundant intercorrelations and overlaps in his responses and condense the whole questionnaire into a much smaller set of variables that he can then name and analyze in whichever way he sees fit.
The Tin Man wrote:
if this dude really isn't sure how to conceptualize his variables because he's working from the ground up and doesn't know what kind of framework is going to emerge from his data then the best thing he can do is just isolate and treat each statement as a separate dependent variable, measured on a likert scale for manageable quantitative analysis. once he's got his dataset, he can run a principle components or factor analysis to eliminate the redundant intercorrelations and overlaps in his responses and condense the whole questionnaire into a much smaller set of variables that he can then name and analyze in whichever way he sees fit.
That is definitely one for Private Eye's pseuds corner. It might actually mean something but what your head be like if you did understand something like that.What kind of brain could come up with that? It's not exactly poetry is it?
AndyPaul wrote:Oldham vs Tranmere on sky tonight![]()
![]()
Featuring for Oldham - Aiden White, Warren Feeney and Paul Dickov.
Featuring for Tranmere - Tony Warner, Andy Robinson and Enoch!
Its going to be shitBut I suppose interesting to see White in action due to out inability to replace Tony Dorigo
Blackwhite wrote:AndyPaul wrote:Oldham vs Tranmere on sky tonight![]()
![]()
Featuring for Oldham - Aiden White, Warren Feeney and Paul Dickov.
Featuring for Tranmere - Tony Warner, Andy Robinson and Enoch!
Its going to be shitBut I suppose interesting to see White in action due to out inability to replace Tony Dorigo
How was it fella? Was out but fucking hell I bet it was greatDickov still looking annoyingly buyable?
AndyPaul wrote:I'll be honest I can barely build up the enthusiasm for champions league and premiership games on TV so there was no fucking way I was gonna waste my energy changing the channel to watch that shite.
gazurtoids wrote:Gus Poyet can get fucked.
MightyWhite wrote:gazurtoids wrote:Gus Poyet can get fucked.
Bloody good manager though.