SPSS for Beginners Part 2: Frequency Counts and Descriptive Statistics

RStatsInstitute Author

05/31/2016 Added

17709 Plays

Description

Introduction to SPSS: Frequency Counts and Descriptive Statistics

Searchable Transcript

Search:

[00:00:01.258]In this video, I'm going to show you
[00:00:02.770]how to do some simple analysis in SPSs.
[00:00:05.243]Now, I already have some data popped in here
[00:00:07.750]but I haven't explained what they are.
[00:00:09.527]Before I do, pause the video and enter these
[00:00:11.605]numbers into your own SPSS spreadsheet.
[00:00:14.870]Now, these two variables represent gender and height.
[00:00:19.379]Variable one represents people's gender information
[00:00:22.375]and variable two represents people's height information.
[00:00:26.043]Now, instead of having to remember this,
[00:00:28.528]let's go into variable view right now,
[00:00:30.629]and change the names of these variables.
[00:00:33.427]We'll call variable One Gender.
[00:00:35.946]We'll call variable Two Height.
[00:00:39.506]Now, the data for people's heights are in inches.
[00:00:42.925]So, we're getting things kind of between 60 and 70 inches,
[00:00:46.512]which is about five to six feet.
[00:00:49.237]The data for people's genders aren't quite as clear.
[00:00:54.135]I'm not typing in male and female,
[00:00:56.427]even though I could using a string variable.
[00:00:59.050]Because male and female, those aren't numbers
[00:01:02.045]and SPSS needs numbers to analyse.
[00:01:04.821]Instead, I'm using a coding scheme of ones and twos.
[00:01:08.548]Where one represents female and two represents male.
[00:01:12.367]Now, this is an arbitrary code.
[00:01:14.341]I could use zeros and ones or fives
[00:01:16.883]and 10s or any two numbers at all.
[00:01:19.461]But, ones and twos are simple, so I'm using those.
[00:01:22.061]Now, a quick mention on the scales of measurement.
[00:01:27.101]Gender is an excellent example of the use of a nominal scale
[00:01:30.815]because the groups are qualitatively and not quantitatively
[00:01:34.461]different so I can change my settings to reflect that.
[00:01:37.816]I'll move over to measure
[00:01:39.653]and change it from Scale to Nominal.
[00:01:42.847]Height, on the other hand, is a quantitative variable
[00:01:46.552]with fixed integral differences between scores
[00:01:48.757]and a meaningful zero so we can leave this
[00:01:51.196]variable as it is set to scale because its ratio.
[00:01:54.469]Now, we're almost set to start analyzing different
[00:01:58.138]aspects of these number but there's
[00:02:00.657]one last thing we can do.
[00:02:04.660]For gender, it can be a little bit confusing to remember,
[00:02:09.221]well which one was male, which one was female?
[00:02:12.901]Ones and twos, I can't quite remember.
[00:02:15.583]So, we can assign value labels to those things.
[00:02:18.756]I go back to variable view and under values.
[00:02:23.303]I can tell SPSS what each value means.
[00:02:27.445]So, I type in that a value of one means female,
[00:02:33.757]it knows that ones are female.
[00:02:35.518]Same thing for twos and male, pop those in there, hit okay.
[00:02:41.382]Now, if I go back to data view, I see all those ones
[00:02:44.901]and twos have been replaced by females and males.
[00:02:48.118]If I want to turn that off and see the numbers again,
[00:02:51.496]there's a little button up here on the toolbar called
[00:02:55.095]Value Labels just click that, they're back to numbers.
[00:02:58.404]Likewise, I can activate it or
[00:03:00.285]reactivate it under the view menu.
[00:03:02.769]Just check this box that says Value Labels
[00:03:05.567]and they'll go back to the males and females.
[00:03:09.213]I think it's more useful to leave
[00:03:11.024]it on so I'm just going to do that.
[00:03:13.044]Now, I think we're ready to analyse these data.
[00:03:16.852]So, probably one of the most simple things you can do
[00:03:19.510]is just count up how often things occur.
[00:03:22.367]It seems easy to do by hand for this data set
[00:03:25.710]but remember if you're dealing with larger data sets
[00:03:27.869]counting things up by hand can get a little tedious.
[00:03:30.993]So, let's have SPSS do it for us.
[00:03:33.048]So, up at the top, go to analyze.
[00:03:36.066]Analyze in SPSS is going to be your very close friend.
[00:03:39.990]Probably the menu you use the most often
[00:03:42.069]and all of the analysis options are in there.
[00:03:44.623]So, under Descriptive Statistics, go to Frequencies.
[00:03:50.032]Here, we can see how often or how
[00:03:52.629]frequent different things occur.
[00:03:54.649]So, not this window pops up.
[00:03:58.096]You're going to see this type of window a lot in SPSS.
[00:04:01.184]Basically, all the different variables
[00:04:04.388]we have are on the left and all the different
[00:04:07.848]variables we want to analyze are on the right.
[00:04:10.912]So you can decide which variables you want to analyze
[00:04:12.851]just by highlighting them and then moving them
[00:04:15.765]over with this little arrow box.
[00:04:17.983]You can move them back if you decide you don't want them.
[00:04:20.108]You can also just kind of grab a hold
[00:04:21.989]of it and then drag it over, whatever you want to analyze.
[00:04:24.787]So, decide whatever you want to analyze.
[00:04:26.702]Gender's probably the best for right now
[00:04:28.549]and when you're ready, just click okay.
[00:04:30.661]Now, your output window's going to pop up here.
[00:04:34.028]I was saying this in the last video,
[00:04:36.268]but one of the problems with SPSs is that
[00:04:39.764]it often gives you a lot more information than you want.
[00:04:42.584]So it's a bit of a trick looking
[00:04:44.407]for the relevant information.
[00:04:46.242]Also the outputs aren't exactly pretty.
[00:04:49.190]So we need to kind of figure out what it's showing us here.
[00:04:52.878]And, luckily, this is a pretty simple table.
[00:04:56.205]It's telling us, first off, this little table,
[00:04:57.934]the N is saying we have 10 valid scores, none are missing.
[00:05:01.882]Basically we don't have any empty cells in there.
[00:05:04.564]This table's probably the more telling one.
[00:05:09.023]It's telling us, on the left are different groups.
[00:05:13.365]We have female and male and in this frequency column,
[00:05:17.323]how often they're occurring.
[00:05:18.821]So, we have six females, we have four males.
[00:05:21.085]We have 10 males total, now there's percent.
[00:05:24.638]Which kind of just convert that to percentages.
[00:05:26.879]60, 40 at 100 percent.
[00:05:28.782]Valid percent if we had a few missing cases it would
[00:05:31.766]divide it differently over the total number
[00:05:34.367]of available cells instead of all cells.
[00:05:37.048]And then cumulative percent which is saying 60 percent
[00:05:40.405]of the people are female and 100 percent
[00:05:43.027]are either female or male.
[00:05:45.443]You know, we really don't need those things.
[00:05:47.904]I think most of the relevant information
[00:05:49.645]is just frequency, right there.
[00:05:53.942]So, we could do the same thing for Height.
[00:05:56.647]We could create a another frequency table for Height.
[00:05:59.050]But, that would tell us how many people are 60 inches tall,
[00:06:02.510]how many people are 61 inches tall,
[00:06:05.354]how many are 62 inches tall, et cetera.
[00:06:08.106]That's probably not going to be very informative to us.
[00:06:11.856]I think a better way to convey the information,
[00:06:14.409]or to convey the gist of how tall people are
[00:06:17.347]might be to calculate the average
[00:06:19.575]of their heights, so let's try that.
[00:06:21.793]And just FYI, you can run analysis from the
[00:06:24.812]output window, the same menu is there.
[00:06:26.948]Just for simplicity or just for stability, I'm going
[00:06:28.992]to go back to the data view of the spreadsheet.
[00:06:32.753]Let's go to analyze.
[00:06:35.038]And, under descriptive stats, go to Descriptives.
[00:06:41.428]We get this same basic window again.
[00:06:43.377]All the variables we have on the left.
[00:06:45.340]All the variables we want to analyze on the right.
[00:06:47.522]Let's just pick Height this time and move it over.
[00:06:50.964]And, before we hit okay, I want to see what different
[00:06:53.264]options we have so click on this button in the upper right.
[00:06:56.305]These are all the different types
[00:06:58.244]and descriptives sets it can calculate for us.
[00:07:00.358]Mean, Standard Deviation, Minimum and Maxinum,
[00:07:02.865]those are pretty standard things.
[00:07:04.723]If we wanted, we could get the Variants and the Range
[00:07:07.277]and the Standard Error of the Mean.
[00:07:09.889]An, the Sum is good too, I'm going to click on the Sum.
[00:07:12.165]There are more options on here like
[00:07:13.802]what type of Distribution we're dealing with.
[00:07:16.170]Display order if we had multiple variables we're
[00:07:18.893]analyzing at once, and actually we can analyze
[00:07:21.738]as many variables as we want at the same time,
[00:07:23.653]but we only have one want to do right now.
[00:07:25.894]We could speficy here what order they appear in.
[00:07:29.052]I really just mostly leave this a variable list.
[00:07:32.094]So, just hit continue and whenever
[00:07:34.439]you're ready, just click okay.
[00:07:38.329]That's what's going to pop up slightly below our last table.
[00:07:43.597]So we see here is our variable, height.
[00:07:47.247]How many scores are in that variable, we have 10.
[00:07:50.068]The minimum score is 61.
[00:07:51.960]That's the shortest person in our group.
[00:07:53.771]80 was the highest score.
[00:07:55.698]692 inches, that would be how long everyone would
[00:07:59.414]be if we kind of stacked them up on top of each other.
[00:08:02.617]But I think this Mean is the most important thing.
[00:08:04.952]So the average height for our 10 people is 69.2 inches.
[00:08:08.829]With a Standard Deviation of about
[00:08:10.710]six point two seven inches.
[00:08:12.939]So we can do now is we can start
[00:08:14.936]to break down Height across categories.
[00:08:17.816]We can compare height of different
[00:08:21.767]people from different genders.
[00:08:24.150]So, the easiest way to do that would be going to analyze.
[00:08:28.133]Instead of Descriptive Stats,
[00:08:30.048]let's go to Compare Means and pick that first option.
[00:08:33.229]Means, now here we need to specify, basically
[00:08:37.855]what our Dependent Variable is
[00:08:40.509]and our Independent Variable.
[00:08:44.551]You know, we didn't really assign people to Condition.
[00:08:48.452]Gender'd be what's called a Quasi Independent Variable.
[00:08:51.866]Let's pretend it's the Independent Variable
[00:08:54.211]because we want to break down the
[00:08:55.848]different categories, the different groups.
[00:08:57.566]Well that's Gender, and the Dependent List,
[00:09:00.329]the things we want to measure, do calculations on,
[00:09:03.604]that'd be Height, the inches, so click on that.
[00:09:07.528]And, I think we should be good to go.
[00:09:09.954]We could see what's under Options.
[00:09:12.183]Ah, a lot of stuff, we don't need that.
[00:09:14.668]What you're going to do is click okay.
[00:09:18.787]Now, once again, we're more than we
[00:09:21.320]really need from SPSS.
[00:09:22.876]This table, I don't really see what it's used for.
[00:09:26.174]This second table is, I think,
[00:09:27.938]the one we really want to see.
[00:09:29.760]This is where we see how tall females are compared to males.
[00:09:32.791]We have females, the average height about 68.3 inches.
[00:09:37.103]And there's six females contributing to that Mean.
[00:09:40.604]Standard Deviation about six point eight nine.
[00:09:42.660]Males, about 70.5 inches, four males going
[00:09:45.423]into that Standard Deviation of about five point nine one.
[00:09:48.117]So, it's telling us, that on average, males are
[00:09:50.520]a little bit taller than females.
[00:09:52.981]We also have the totals 69.2 inches,
[00:09:56.023]that's how tall people are on average.
[00:09:58.704]All 10 people, with a Standard Deviations
[00:10:01.305]of about six point two seven.
[00:10:03.244]That's the same number we got up here from doing
[00:10:05.264]Descriptive for all those people at once, 69.2.
[00:10:08.399]So overall Frequency counts and Descriptor Stats
[00:10:11.324]are a good way to take a peek at
[00:10:13.902]potential trends in your data.
[00:10:15.597]And, SPSS made it extremely easy to do so.

The screen size you are trying to search captions on is too small!

You can always jump over to MediaHub and check it out there.

Comments

0 Comments

SPSS for Beginners Part 2: Frequency Counts and Descriptive Statistics

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments