SPSS FILE TYPES 2.

Transform Tab

- .sav yung default file type, siya yung mismong Compute Variables
data
- Transform or kaya magcompute (mga mean
- .spv kapag output
ganon using spss) new variables galling sa mga
existing variables

TYPES OF FILES (na hindi SPSS made) Recode Variables

csv – comma separated text file - Group continuous variables into intervals
- Redefine an existing numeric code
xls – excel files
- Convert string values into numeric codes
.dta – stata data sets
Visual Binning

- Ginugroup into categorical tas papakita ng

When importing the types of files from above, you can parang visual representation nung variable (??)
choose between…
3. Utilities Tab
a. Delimiter - may ginagamit na character to
separate values in a line Manipulating Variables
b. Fixed width – may non-overlapping ranges of
- Defining Variables Sets: Utilities > Define
columns to separate variables; yung may line na
Variable Sets tapos if okay na pindutin 2 venn
inuurong
diagram tas mamili ng variable na ipapakita.
Functions in SPSS Useful ito kapag gusto mo temporarily ihide
ilang variables if nahihilo ka sa dami ng
1. Data Tab variables. If di mahanap yung venn diagram,
Merge Cases Utilities > Use Variable Sets. Ginagamit para
masala yung variables.
- Madali if 2 files same sila properties yung
variables (like ordinal ganon) Corollary: key variable yung parang variable na
pagbabasehan para ma-sort mo. Parang eto yung
Select Cases magpapa-unique dun sa data set mo para mamatch
- Useful for filtering out cases TEMPORARILY OR mo yung nasa dataset 1 at dun sa iincoporate mo pa
PERMANENTLY lang na dataset. Kunwari data ng upd students,
possible key variable is student number.
Exploring Cases
- If may key variable, both data sets will be sorted
- Identifying Duplicates: yes yun na yon pero then tsaka imamatch from top to bottom.
pipili ng key indicators Hanggang mameet ng mismatch, will continue
- Sorting within a match: 2nd level of validation yung pagmamatch.
- Sort Cases: sort accdg to numerical, - Allowed ang many to one (marami from orig, isa
chronological or alphabetical. To do this, Data > from supplemental) pero ang one to many hindi
Sort Cases. Dito masesave mo yung sorted file (isa from orig to many supplemental dataset).
into a new SPSS data set. Pag yung latter, yung unang match lang
marerecord.
Adding Variables
- If meron yung orig data set tas wala sa
- Kapag walang designated key variable, matik na supplmental, mabblank lang yung cell na yun.
parang minamatch lang ni SPSS kapag nag-add Pero pag wala sa orig data set tas meron sa
ka ng variables, from top top bottom. supplemental, di marerecord.
- To do this, Data > Merge Files > Add Variables
 4. Analyze Tab - for figuring out relationships between 2
variables and other types of tables
Descriptive Statistics

a. Frequencies:
- if need ng frequency dist per variable
- Sum
- Central tendencies: 3 Ms,
- Position: percentiles
- Dispersion: Standard dev, range, min, max
- Tailedness: Kurtosis, skewness
- Bar, pie, histogram
- Pwede di pakita frequency tables untick lang
button sa baba (usually ganto if continuous
yung variable)
- More generic way for fractiles: cut points. If 99, - Chi-Square Test of Independence: appropriate
percentiles yun. If 10, deciles and so on. in testing the existence of a relationship
- Standard Error of the sample mean: how diff between 2 categorical variables. If di tinick yung
the pop mean is likely to be from the sample suppress tables, maglalabas si spss ng
mean. Iba toh sa S.D. kasi SD pertains to how contingency table. Naglalabas din ng continuity
dispersed the data is. So itong standard erro, correction paa ma-fill gaps between “very
yung possible error na ganto kalayo yung discrete” values.
sample mean mo sa population

b. Descriptives
- Mean, Sum
- Dispersion: Standard dev, range, min, max, S.E.
mean
- Tailedness: Kurtosis, skewness
- Can generate z-scores (yung tick na
standardized values). Masesave silaas new
variables like, “z[var name]”.
- expected counts

o Explore
- Boxplot, stem and leaf display
- Summary measures c. Tables
- Normality plots and results of tests for o Custom tables: most flexible way to construct
normality summary or frequency tables
- Has an option to include a factor
variable for “deep diving”

o Crosstabs
- Generates 2-way or contingency tables
- Generates expected frequency tables
- Generates results for diff tests for
categorical data/2-way tables (chi-
square)
- May chi-square at correlation
o Multiple Response Sets: For specifying a gamitin. Tas grp 1 must be >= cutpoint
collection of variables that defines a single and grp 2 must be < cutpoint/
multiple response item. Yung defined sets nito - Nilalabas neto yung 2 possible na
ginagamit in custom tables to create summaries pwedeng equal or unequal variances
for multiple response items properly If - If no prior info whether equal or not
variances, Levene’s Test ginagamit.
dichotomized, parang binary lang possible
responses. Pag categorical, may designated na
o Paired Sampled T-test
pare-parehas at fixed na possible values (e.g. - Assuming samples before and after
pag educ attainment, kinder to doctoral ganun intervention program came from a normal
lang). If nakapagset ka ng ganto, magaappear population
siya if gagawa ka custom tables. Ginagamit if - Dapat may variables ka na pwede ipair tas
separate columns in the data
may multiple responses ka sa isang variable.
- Yung D niyan is D = Var1 – Var2
- Computed dito si Pearson’s Moment
Correlation Coefficient kasi (1) paired
measures are intuitively correlated, (2) it
assumes sample galing BVN, and (3) PMCC
is derived based on same distn assump on
(2).
- Yung tatlong dulo yung result niyan

o One-way ANOVA
- If galing sa diff pop but same variances
- Always assumed equal variances dito
- One-way meaning single factor
- Bawal continuous as grouping variable
d. Compare Means - Yung test variable yung dependent kasi it is
viewed as the one who “responds” sa diff
- testing inferences about a pop mean levels of factor or treatments
o One-Sample T-test - Sa Options makikita yung stats
- Ginagamit parang yung sa 115 na mga - Post Hoc: to compare pairs of grps and
testing hypothesis ganern detect kung aling pair ang significantly diff
- Yung test value don yung hypothesized means. Appropriate only if may sufficient
value ng parameter of interest. evidence to conclude alternative hypo of
ANOVA.
- When working with multiple datasets,
- Significance level: max error na inaallow sa
need tignan yung active data set sa taas LAHAAAT ng pairwise comparisons
nung bago yung output na tables. - LSD (Least Significant Difference): a post hoc
- Sig always refer to the P-value test na nagcconduct ng all possible pairwise
comparisons.
o Two-Independent Sample T-test - (wweee) ginagamit if maraming roups na
- can use for equal or unequal population gustong itest
variances - naiiba toh kasi tinetest at least 1 grp has
- Confounding variable: unaccounted naiibang mean tas p[agkareject null;,
initiate post hoc test. tsaka hahanapin
variable na nagccause ng changes sa
(wweee)
isang variable (kung sa correlation may
x at y, siya si z). e. Correlate
- Sa Define Groups, group 1 yung - Pearson's: both variables must be at least interval
population parameter myu sub 1 tas - Spearman & Kendall’s Tau: at least ordinal
yung group 2 naman yung pop - Correlation is a measure of linear relationship
parameter my sub 2 na inestimate sa - Partial Correlation: relationship between 2 after
hypothesis. In other words, group 2 considering other variables affecting the 2.
yung value na tinetest sa hypothesis. - Correlation of the variable to itself is 1.
- If at least ordinal, pwede non-
f. Regression
dichotomous variable tas Cut Point
- Simple linear pwede para idescribe how a Kapag nagssummary tables, always activate weights by
particular variable reacts to changes to a Data > Weight Cases. Incorrect ang summarized data
possible relevant factor. pwede rin to predict a kung wala.
target variable. Yung pinepredict dito yung
If by-group processing, analysis dapat nonoverlapping
dependent variable tas predictor
groups of cases like computing sample mean by male
- y = xB1 + B0 + Ep where
and female ganern. to do this, pwede Data > Split File.
y is dependent Nirerequire neto na sorted ang data by criteria.

x is independent Sa slides Compare Groups, parang hindi nagweight

cases. Nagsplit file tas nilagay yung pangsplit which is
B1 is slope parameter
sex. Tas wala mismong output. need magdescriptives
B0 is intercept parameter tas kinuha mean at s.d. nung ibang variable.

Ep is random error where ep follows Normal

with mean 0 and variance sigma-squared.
g. Exporting
Meaning variance is constant, does not depend
on the value of the predictor - File > Export

- sa class natin both variables are continuous

- Pag tinick yung Normal probability plot,
h. Syntax
magegenerate yon ng Normal Quantile-Quantile
plot for residuals.Dinedetermine neto whether - SPSS language; often called command
or not yung error term follows a normal - generated every time may ineexecute na
distribution procedure
- Sa Save, pag tinick yung Unstandardized sa - Five command types:
Predicted values at Residuals, maglalabas si spss - File Definition - input data into SPSS
ng new variable for predicted values nila at (import, match files, get, data list)
residual bawat observation - Input Program - more specialized lang
- Yunh R-squred or coefficient of determination, nung una (record type, repeating data,
yan yung percent of variation sa dependent na end case)
ineexplain ng independent - Transformation - modify data (compute,
- ANOVA in regression naman yung overall recode)
significance test for a model, if yung isa sa mga - Procedure - analyzing data
predictors ba can predict yung response tas null (descriptives, crosstabs) example nito is
niyan none of the predictors are significant yung Save kasi it causes data to read
- P-value naman for test of significance of slope - Utility - variety of chores, to add
and inter (whether or not equal to 0) comments, define new file, define
macro
 - Split File: hinahati data set into separate groups,
pwedeng accdg to a variable or nakayos na first
ganto karami obs ganon. Compare Groups, side
REVIEW SESSION NI SIR PAJARON by side for easier comparison. Organize oput
put by greoups (separate other term),
● Statistics-oriented magkahiwalay.
- pag wala sa range yung value sa pinanggawa na - Correlate: pwede lagay konting stat. May
equation, di na valid kendall’s tau-b, spearman at pearson.
● SPSS-Oriented - Partial correlation: ginagamit if may
- if nagt-test ka, magkasama na CI at hypothesis confounding variable
(?) - Canonical correlation: mga may hidden variable
- Sa pagsesave ng output, pwede rin magamit sa - Regression:
word. Export gagawin instead of save. - Sa graph
- sa data view naglalagay ng data, process na - yung ANOVA: nagaappear regression kasi
naglalagay extra importas related concept sila.
- Variable view sa variable properties, can define - Sa regression, binbreakdwon yung mga
new variables, usually used to ready sa mga variables. titignan aling variable significant if
ipapasok na data nagreject sa anova
- numeric for computations - One way anova i f may groups na naiiba
- Formatted numeric, special na numeric mga - magpoposthoc test in testing many means if
dollar sign or scientific notation, pwedeng date nagreject sa anova
- if magpapasok ng data na may specific format, - Aniva kasi nagagamit laaht ng data unlike other
need marecognize ng computer yung format. na di nagagamit lahat ng data
- width: number of characters - SUmmary tables
- Decimals: # of decimals, 0 always when
character
- Label: kung ano ibig sabihin ng name. walang
limitation dito unlike name na bawal space
ganon. eto gagamitin na label
- Values: value labels, yung 1 is yes 0 is no ganon
- Missing: to separate inispecify na value (like 000
is missing, na hihiwalay obs na ganyan in
analyses)
- Pag nagpaste ka lang, lalabas syntax form. so iuf
gusto mo ilabas output wag paste.
- .sav yung default file type, siya yung mismong
data
- .spv kapag output
- pag galing ibang software, import na
- Weight cases: allow u to consider enter
summary data. tas kunwari weight by
frequency.
- Select Cases: when u want something to be
done for the next remaining cases or ganon.
Pwedeng strike out lang yung di gagamitin until
idisable so temporary. Pwede rin gumaa ng new
fille na andon selected cases. pwede rin idelete
yung mga di napili. Pwede rin toh if pipili mga
nasatisfy lang yung condition. May random