4 independent variables, no dependent
Locked
 
|
Hi everyone,
I would appreciate it a lot if someone could help me and my team figure this one out. The stress is on ASPECT III. In the dataset we have 4 scale variables and we are asked to figure out which one of them is the most important in determining the success of motion picture. The respondents were asked to determine the success based on those four factors, CAST, SCRIPT, CRITICS, AUDIENCE. Our initial idea was to compare means and see which one of them scored highest, as scales are from 1-10. However, we think this is too simpe for 3rd question, as question difficulty increases from first, being the easiest, to last, being the hardest. Can someone help us and suggest some type of test or are we simpy right and only need descriptive statistics? I hope pictures can help you in understanding the issue. Please type me for more if you need any further data. <http://spssx-discussion.1045642.n5.nabble.com/file/t341546/Unbenannt.png> <http://spssx-discussion.1045642.n5.nabble.com/file/t341546/Unbe1nannt.png> -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
Re: 4 independent variables, no dependent
|
Looks like a coursework assignment to me, but as well as comparing means you
could also compare the frequency distributions, especially the percentages on points 8-10. If the distribution is skewed you might be better comparing the medians. Is there a dependent variable anywhere? John F Hall [Retired academic survey researcher] IBM-SPSS Academic Author 9900074 Email: [hidden email] Website: http://surveyresearch.weebly.com / SPSS course: http://surveyresearch.weebly.com/1-survey-analysis-workshop-spss.html Research: http://surveyresearch.weebly.com/3-subjective-social-indicators-quality-of-l ife.html -----Original Message----- From: SPSSX(r) Discussion <[hidden email]> On Behalf Of vasom Sent: 14 December 2018 14:57 To: [hidden email] Subject: 4 independent variables, no dependent Hi everyone, I would appreciate it a lot if someone could help me and my team figure this one out. The stress is on ASPECT III. In the dataset we have 4 scale variables and we are asked to figure out which one of them is the most important in determining the success of motion picture. The respondents were asked to determine the success based on those four factors, CAST, SCRIPT, CRITICS, AUDIENCE. Our initial idea was to compare means and see which one of them scored highest, as scales are from 1-10. However, we think this is too simpe for 3rd question, as question difficulty increases from first, being the easiest, to last, being the hardest. Can someone help us and suggest some type of test or are we simpy right and only need descriptive statistics? I hope pictures can help you in understanding the issue. Please type me for more if you need any further data. <http://spssx-discussion.1045642.n5.nabble.com/file/t341546/Unbenannt.png> <http://spssx-discussion.1045642.n5.nabble.com/file/t341546/Unbe1nannt.png> -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
Dear John,
Thanks for fast reply. Yes, it is coursework assignment, sending you file right away. Hopefully you could help us! Thank You in advance Sir, Vaso M. -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
Re: 4 independent variables, no dependent
|
|
This list is not here to do your (? MBA ?) homework for you, but to help you
to learn and execute SPSS. Please make your course tutor aware of this. John F Hall MA (Cantab) Dip Ed (Dunelm) [Retired academic survey researcher] Email: [hidden email] Website: Journeys in Survey Research Course: Survey Analysis Workshop (SPSS) Research: Subjective Social Indicators (Quality of Life) -----Original Message----- From: SPSSX(r) Discussion <[hidden email]> On Behalf Of vasom Sent: 15 December 2018 17:48 To: [hidden email] Subject: Re: 4 independent variables, no dependent Dear John, Thanks for fast reply. Yes, it is coursework assignment, sending you file right away. Hopefully you could help us! Thank You in advance Sir, Vaso M. -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
Re: 4 independent variables, no dependent
|
|
In reply to this post by vasom
John did not ask for you to send him any file.
Tempted to delete this thread. ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
Re: 4 independent variables, no dependent
|
|
In reply to this post by vasom
We don't do homework, but I have sympathy with the Poster because this is
a really crappy assignment - and he might pass along this criticism. Someone
has concocted a problem with odds and ends of real jargon, but the pieces do
not add up to an intelligible problem. (There is always this problem in trying to
construct "realistic" homework which modifies a standard design.... This attempt
failed.)
Nominally, the data come from an audience (of the movie "The Residuals"), since
their responses are supposed to relate to one movie. But, then, they are asked
generically about what matters to the "success" of a movie. Okay, you will get some
amateur opinions, which will be influenced in varying degrees by what is right or
wrong about they one they just saw.
Basic design rules.
(a) If you want to know the assets of THIS movie, ask about THIS movie.
(b) If you want to set up a criterion, be explicit and narrow so that you do not
get idiosyncratic readings. Is "success" the gross income ("advertising" wins),
good movie, or net income? A lot of the best movies never make much money.
Viewers should be asked about the strengths/weaknesses of what they just saw.
Explicitly. "Actors are important" does not say if these were great or terrible, by
being known or not, or by being good or not.
Analysis.
The question of "which variable is most important" among close competitors is
naturally attractive, but the statistical answers are seldom very useful when you
can't set up orthogonal contrasts. The big two bugaboos are "confounding" and
"ranges of variation". When two variables overlap, they are confounded. When
you consider only a narrow range of variation for some variable, then its influence
will be much smaller than when you consider a wide range. (see: "nature vs. nurture".)
Hope this helps.
--
Rich Ulrich
From: SPSSX(r) Discussion <[hidden email]> on behalf of vasom <[hidden email]>
Sent: Friday, December 14, 2018 8:56 AM To: [hidden email] Subject: 4 independent variables, no dependent Hi everyone,
I would appreciate it a lot if someone could help me and my team figure this one out. The stress is on ASPECT III. In the dataset we have 4 scale variables and we are asked to figure out which one of them is the most important in determining the success of motion picture. The respondents were asked to determine the success based on those four factors, CAST, SCRIPT, CRITICS, AUDIENCE. Our initial idea was to compare means and see which one of them scored highest, as scales are from 1-10. However, we think this is too simpe for 3rd question, as question difficulty increases from first, being the easiest, to last, being the hardest. Can someone help us and suggest some type of test or are we simpy right and only need descriptive statistics? I hope pictures can help you in understanding the issue. Please type me for more if you need any further data. <http://spssx-discussion.1045642.n5.nabble.com/file/t341546/Unbenannt.png> <http://spssx-discussion.1045642.n5.nabble.com/file/t341546/Unbe1nannt.png> -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
Re: 4 independent variables, no dependent
|
|
I fully agree. I looked at the assignment and, out of curiosity, asked Vaso [off-list] to send me the SPSS *.sav file. Data could be artificial (N=120, no data missing for any variable). OK as an albeit primitive exercise for basic operations to look at distributions, compare sex and age groups, or association between variables, but that's all. There is no dependent variable and the sample is far too small. There are plenty of SPSS *.sav files for major surveys available from archives, many of which also have open-source reports: why don't tutors use these for both primary and secondary analysis? Time and again the culprits are MBA courses on which students tick off weekly exercises, but rarely if ever learn anything about data management and analysis, let alone SPSS. John F Hall MA (Cantab) Dip Ed (Dunelm) [Retired academic survey researcher] Email: [hidden email] Website: Journeys in Survey Research Course: Survey Analysis Workshop (SPSS) From: SPSSX(r) Discussion <[hidden email]> On Behalf Of Rich Ulrich We don't do homework, but I have sympathy with the Poster because this is a really crappy assignment - and he might pass along this criticism. Someone has concocted a problem with odds and ends of real jargon, but the pieces do not add up to an intelligible problem. (There is always this problem in trying to construct "realistic" homework which modifies a standard design.... This attempt failed.) Nominally, the data come from an audience (of the movie "The Residuals"), since their responses are supposed to relate to one movie. But, then, they are asked generically about what matters to the "success" of a movie. Okay, you will get some amateur opinions, which will be influenced in varying degrees by what is right or wrong about they one they just saw. Basic design rules. (a) If you want to know the assets of THIS movie, ask about THIS movie. (b) If you want to set up a criterion, be explicit and narrow so that you do not get idiosyncratic readings. Is "success" the gross income ("advertising" wins), good movie, or net income? A lot of the best movies never make much money. Viewers should be asked about the strengths/weaknesses of what they just saw. Explicitly. "Actors are important" does not say if these were great or terrible, by being known or not, or by being good or not. Analysis. The question of "which variable is most important" among close competitors is naturally attractive, but the statistical answers are seldom very useful when you can't set up orthogonal contrasts. The big two bugaboos are "confounding" and "ranges of variation". When two variables overlap, they are confounded. When you consider only a narrow range of variation for some variable, then its influence will be much smaller than when you consider a wide range. (see: "nature vs. nurture".) Hope this helps. -- Rich Ulrich From: SPSSX(r) Discussion <[hidden email]> on behalf of vasom <[hidden email]> Hi everyone, ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
| Free forum by Nabble | Edit this page |