SlideShare uma empresa Scribd logo
1 de 23
Chapter 4
Changing the Way We Do: Hypothesis
Testing, Power, Effect Size, and
Other Misunderstood Issues
Annisa Fitri Irwan
Della Oferischa
Musfera Nara Vadia
Rezky Jafri
There are a few important steps that researchers can take to make the
statistical results from their studies meaningful and useful. They are:
 Perform a power analysis before undertaking a study in order to
determine the number of participants that should be included in a
study to ensure an adequate level of power (power should be higher
than .50 and ideally .80). Briefly put, power is the probability that you
will find differences between groups or relationships among variables if
they actually do exist.
 Never set an alpha level lower than .05 and try to set it higher, to
.10 if at all acceptable to the research community one is working in.
 Report effect sizes and their interpretation.
 Report confidence intervals.
1. Null Hypothesis Significant Tests
 Null hypothesis (Ho) is there is no difference
between groups or that there is no relationship
between variables.
 Once we reject null hypothesis we will be able
to accept alternative hypothesis (Ha).
Example:
 HR: Will 15 minutes of practice of meaningful drills result in
more accurate grammar scores than 15 minutes of telling a story
where the grammar in question must be used?
 H0 : There is no [statistical] difference between a group which
practices grammar using explicit meaningful drills for 15 minutes
each day and a group which uses grammar implicitly by telling
stories where the grammar is needed for 15 minutes each day.
 Ha: There is a [statistical] difference between the explicit and
implicit group.
 Since only two groups are being compared, a t-test can be used. The
t-test statistic is calculated based on three pieces of information.:
the mean scores of the groups, their variances, and the size of each
group (the sample size).
 In NHST process, we should have already decided on a cut-off level
that we will use to consider the results of the statistic test extreme.
This is called alpha level or significant level.
 Baayen (2008) “if the p-value is lower than alpha level we set, we
reject the null hypothesis and accept the alternative hypothesis that
there is a difference between the two groups (it does not necessarily
mean the alternative hypothesis is correct)”
 P-value: the probability of finding a [insert statistic name here] this
large or larger if the null hypothesis were true is [insert p-value].
One-Tailed versus Two-Tailed Tests of
Hypothesis
 if the only thing we care about is just one of the possibilities, then we can
use a one-tailed test. A one-tailed or directional test of a hypothesis looks
only at one end of the distribution. A one-tailed test will have more power to
find differences, because it can allocate the entire alpha level to one side of
the distribution and not have to split it up between both ends.
 Two-tailed hypothesis is examining two possibilities of groups. The
hypothesis could go in either direction. For example: in our null hypothesis,
we would be examining both the possibility that the explicit group was better
than the implicit group and the possibility that the explicit group was worse
than the implicit group.
Outcomes of Statistical Testing
True situation in the populationOutcomeobservedinstudy
No effect exist
No effect exist Effect exist
Correct
situation
(probability= I-
α)
Type II error
(probability=ß)
Effect exist Type I error
(probability=α)
Correct
situation
(probability= I-
ß)
Power
Type I error (being
overeager): concluding there
is a relationship when there
is none.
 Set type I error level by
setting alpha (α) level
o Commonly set at α = .05
o Possibility of a Type I error
is thus 5%
Type error II (being overly
cautious): concluding there is
no relationship when there is
one.
 Set type II error level (ß) and then
calculate power (power = I – ß)
o Commonly set at ß = .20, resulting in
power =.80
o Possibility of Type II error is thus 20%
 Avoid low power by:
o Having adequate sample sizes
o Using reliable dependent variable
o controlling for individual differences
o Including pre-test
o Using longer post-test
o Making sure not to violate statistical
assumptions.
Problems with NHST
There are problems with using NHST method of making
conclusions about experiments.
 One is that some authors interpret a low p-value as
indicate of a very strong result. On the other hand,
probably a lower p-value does not make a study more
significant, in the generally accepted sense of being
important.
 On the other hand, the p-value of a study is an index of
group size and the power of study, and this is why a p-
value of .049 and a p-value of .001 are not equivalent,
although both are lower than α=.05
Change the Way I Do Statistics
 Reporting exact p-values (unless they are so small
it would take too much room to report them)
 Talking about “statistical” results instead of
“significant” or “statistically significant” results
 Providing confidence intervals and effect size
whenever possible.
2. Power Analysis
 Power is the probability of detecting a statistical result
when there are in fact differences between groups or
relationships between variables.
 Power often translates into the probability that the test
will lead to a correct conclusion about the null
hypothesis.
What are the theoretical implications if
power is not high?
 If the power of a test is .50, this means that there
is only a 50% chance that a true effect will be
detected. In other words, even though there is in
fact an effect for some treatment, the researcher
has only a 50/50 chance of finding that effect.
What is the optimal level of power?
 Power should be above .50 and would be judged
adequate at .80. a power level of .80 would mean
that four out of five times a real effect in the
population will be found.
 Power levels ought to be calculated before a
study is done, and not after.
Help with calculating power using R
 How to use the arguments for the “pwr” library, how to
calculate effect sizes, and Cohen’s guidelines as to the
magnitude of effect sizes. Cohen meant for these guidelines
to be a help to those who may not know how to start doing
power analyses, but once you have a better idea of effect
sizes you may be able to make your own guidelines about
what constitutes a small, medium, or large effect size for the
particular question you are studying.
 Remember that obtaining a small effect size means that you
think the difference between groups is going to be quite
small.
3. Effect Size
 Effect size is the magnitude of the impact of the independent variable on
the dependent variable.
 An effect size gives the researcher insight into the size of the difference
between groups is important or negligible.
 If the effect size is quite small, then it may make sense to simply discount
the findings as unimportant, even if they are statistical.
 If the effect size is large, then the researcher has found something that it is
important to understand.
 Effect sizes do not change no matter how many participants there are, that
makes effect sizes a valuable piece of information, much more valuable
than the question of whether a statistical test is “significant” or not.
Understanding Effect Size Measures
Huberty (2002) divided effect sizes into
two broad families: group difference
indexes and relationship indexes. Both the
group difference and relationship effect
sizes are ways to provide a standardized
measure of the strength of the effect that
is found.
 A group difference index, or mean difference measure,
has been called the d family of effect sizes by Rosenthal
(1994). The prototypical effect size measure in this family
is Cohen’s d. Cohen’s d measures the difference between
two independent sample means, and expresses how large
the difference is in standard deviations.
 Relationship indexes, also called the r family of effect
sizes, measure how much an independent and dependent
variable vary together or, in other words, the amount of
covariation in the two variables. The more closely the two
variables are related, the higher the effect size.
Calculating Effect Size for Power Analysis
How to determine the effect size to expect?
 The best way to do this is to look at effect sizes from previous in this
area to see what the magnitude of effect sizes has been. If there is
none, the researcher must make an educated guess of the size of the
effect that they will find acceptable, or use Cohen’s effect size.
 Cohen notes that effect size are likely to be small when one is
undertaking research in an area that has been little studied, as
researchers may not know yet what kinds of variables they need to
control for.
 A small effect size is one that is not visible to the naked eye but exists
nevertheless.
 Cohen notes that effect size magnitudes depend on the area.
Calculating Effect Size Summary
 In general, statistical tests which include a categorical variable
that divides the sample into groups, such as the t-test or ANOVA,
use the d family of the effect sizes to measure effect size. The
basic idea of effect sizes in the d family is to look at the
difference between the means of two groups, as in µA-µB.
Table 4.6 Options for Computing Standardizers (the Denominator
for d Family Effect Sizes)
 A The standard deviation of one of the groups, perhaps most
typically the control group
 B The pooled standard deviation of [only the groups] being
compared
 C The pooled standard deviation [of all the groups] in the
design
4. Confidence Intervals
 The confidence interval represents “a range of plausible
values for the corresponding parameter”, whether that
parameter be the true mean, the difference in scores or
whatever.
 The width of the confidence interval indicates the
precision with which the difference can be calculated or,
more precisely, the amount of sampling error.
 If there is a lot of sampling error in a study, then
confidence intervals will be wide, and the statistical
results may not be very good estimates.
Power through Replication and Belief in
the “Law of Small Numbers”
 Tversky and Kahneman (1971) point out that replication studies will ideally
have a larger number of participants than the original study. With a sample
size that is larger than the original, the experimenter will have a better
chance of finding a significant result. Sample sizes play a direct role in the
amount of power that a study has, and also directly affect the p-value of the
test statistic.
 Tversky and Kahneman’s proposed law of small numbers states that “the law of
large numbers applies to small numbers as well”
 In other words, researchers believe that even a small sampling should
represent the whole population well, which leads to an unfounded confidence
in results found with small sample sizes.
Larson-Hall, Jenifer. 2010. A Guide to doing Statistic in Second Language Research
Using SPSS. New York: Routledge

Mais conteúdo relacionado

Mais procurados

Reporting a Factorial ANOVA
Reporting a Factorial ANOVAReporting a Factorial ANOVA
Reporting a Factorial ANOVAKen Plummer
 
Mixed between-within groups ANOVA
Mixed between-within groups ANOVAMixed between-within groups ANOVA
Mixed between-within groups ANOVAMahsa Farahanynia
 
Introduction to Power Analysis
Introduction to Power AnalysisIntroduction to Power Analysis
Introduction to Power AnalysisDaria Bondareva
 
Parametric Statistical tests
Parametric Statistical testsParametric Statistical tests
Parametric Statistical testsSundar B N
 
Writing up your results – apa style guidelines
Writing up your results – apa style guidelinesWriting up your results – apa style guidelines
Writing up your results – apa style guidelinesMarialena Kostouli
 
Confidence intervals
Confidence intervalsConfidence intervals
Confidence intervalsTanay Tandon
 
Reporting a one way ANOVA
Reporting a one way ANOVAReporting a one way ANOVA
Reporting a one way ANOVAAmit Sharma
 
Checking for normality (Normal distribution)
Checking for normality (Normal distribution)Checking for normality (Normal distribution)
Checking for normality (Normal distribution)Rizwan S A
 
Kruskal Wallis test, Friedman test, Spearman Correlation
Kruskal Wallis test, Friedman test, Spearman CorrelationKruskal Wallis test, Friedman test, Spearman Correlation
Kruskal Wallis test, Friedman test, Spearman CorrelationRizwan S A
 
Anova (Analysis of variation)
Anova (Analysis of variation)Anova (Analysis of variation)
Anova (Analysis of variation)Shakeel Rehman
 
Multiple Linear Regression II and ANOVA I
Multiple Linear Regression II and ANOVA IMultiple Linear Regression II and ANOVA I
Multiple Linear Regression II and ANOVA IJames Neill
 

Mais procurados (20)

Reporting a Factorial ANOVA
Reporting a Factorial ANOVAReporting a Factorial ANOVA
Reporting a Factorial ANOVA
 
Mixed between-within groups ANOVA
Mixed between-within groups ANOVAMixed between-within groups ANOVA
Mixed between-within groups ANOVA
 
Introduction to Power Analysis
Introduction to Power AnalysisIntroduction to Power Analysis
Introduction to Power Analysis
 
Repeated Measures ANOVA
Repeated Measures ANOVARepeated Measures ANOVA
Repeated Measures ANOVA
 
Factor analysis
Factor analysisFactor analysis
Factor analysis
 
Parametric Statistical tests
Parametric Statistical testsParametric Statistical tests
Parametric Statistical tests
 
Pearson Correlation
Pearson CorrelationPearson Correlation
Pearson Correlation
 
Writing up your results – apa style guidelines
Writing up your results – apa style guidelinesWriting up your results – apa style guidelines
Writing up your results – apa style guidelines
 
Sample size determination
Sample size determinationSample size determination
Sample size determination
 
Z test
Z testZ test
Z test
 
Confidence intervals
Confidence intervalsConfidence intervals
Confidence intervals
 
Reporting a one way ANOVA
Reporting a one way ANOVAReporting a one way ANOVA
Reporting a one way ANOVA
 
Parametric Test
Parametric TestParametric Test
Parametric Test
 
Posthoc
PosthocPosthoc
Posthoc
 
Checking for normality (Normal distribution)
Checking for normality (Normal distribution)Checking for normality (Normal distribution)
Checking for normality (Normal distribution)
 
Analysis of Variance (ANOVA)
Analysis of Variance (ANOVA)Analysis of Variance (ANOVA)
Analysis of Variance (ANOVA)
 
Kruskal Wallis test, Friedman test, Spearman Correlation
Kruskal Wallis test, Friedman test, Spearman CorrelationKruskal Wallis test, Friedman test, Spearman Correlation
Kruskal Wallis test, Friedman test, Spearman Correlation
 
Effect Size
Effect SizeEffect Size
Effect Size
 
Anova (Analysis of variation)
Anova (Analysis of variation)Anova (Analysis of variation)
Anova (Analysis of variation)
 
Multiple Linear Regression II and ANOVA I
Multiple Linear Regression II and ANOVA IMultiple Linear Regression II and ANOVA I
Multiple Linear Regression II and ANOVA I
 

Destaque

Interim analysis in clinical trials (1)
Interim analysis in clinical trials (1)Interim analysis in clinical trials (1)
Interim analysis in clinical trials (1)ADITYA CHAKRABORTY
 
Review & Hypothesis Testing
Review & Hypothesis TestingReview & Hypothesis Testing
Review & Hypothesis TestingSr Edith Bogue
 
Chapter 11
Chapter 11Chapter 11
Chapter 11bmcfad01
 
Inferential statistics powerpoint
Inferential statistics powerpointInferential statistics powerpoint
Inferential statistics powerpointkellula
 
What's Significant? Hypothesis Testing, Effect Size, Confidence Intervals, & ...
What's Significant? Hypothesis Testing, Effect Size, Confidence Intervals, & ...What's Significant? Hypothesis Testing, Effect Size, Confidence Intervals, & ...
What's Significant? Hypothesis Testing, Effect Size, Confidence Intervals, & ...Pat Barlow
 
Test of hypothesis
Test of hypothesisTest of hypothesis
Test of hypothesisvikramlawand
 
Hypothesis testing ppt final
Hypothesis testing ppt finalHypothesis testing ppt final
Hypothesis testing ppt finalpiyushdhaker
 
Hypothesis testing; z test, t-test. f-test
Hypothesis testing; z test, t-test. f-testHypothesis testing; z test, t-test. f-test
Hypothesis testing; z test, t-test. f-testShakehand with Life
 

Destaque (13)

Abc4
Abc4Abc4
Abc4
 
Interim analysis in clinical trials (1)
Interim analysis in clinical trials (1)Interim analysis in clinical trials (1)
Interim analysis in clinical trials (1)
 
Review & Hypothesis Testing
Review & Hypothesis TestingReview & Hypothesis Testing
Review & Hypothesis Testing
 
Chapter 11
Chapter 11Chapter 11
Chapter 11
 
Testing Hypothesis
Testing HypothesisTesting Hypothesis
Testing Hypothesis
 
Inferential statistics powerpoint
Inferential statistics powerpointInferential statistics powerpoint
Inferential statistics powerpoint
 
Hypothesis
HypothesisHypothesis
Hypothesis
 
What's Significant? Hypothesis Testing, Effect Size, Confidence Intervals, & ...
What's Significant? Hypothesis Testing, Effect Size, Confidence Intervals, & ...What's Significant? Hypothesis Testing, Effect Size, Confidence Intervals, & ...
What's Significant? Hypothesis Testing, Effect Size, Confidence Intervals, & ...
 
P value
P valueP value
P value
 
Hypothesis Testing
Hypothesis TestingHypothesis Testing
Hypothesis Testing
 
Test of hypothesis
Test of hypothesisTest of hypothesis
Test of hypothesis
 
Hypothesis testing ppt final
Hypothesis testing ppt finalHypothesis testing ppt final
Hypothesis testing ppt final
 
Hypothesis testing; z test, t-test. f-test
Hypothesis testing; z test, t-test. f-testHypothesis testing; z test, t-test. f-test
Hypothesis testing; z test, t-test. f-test
 

Semelhante a Hypothesis Testing, Effect Size, and Power Analysis

© 2014 Laureate Education, Inc. Page 1 of 5 Week 4 A.docx
© 2014 Laureate Education, Inc.   Page 1 of 5  Week 4 A.docx© 2014 Laureate Education, Inc.   Page 1 of 5  Week 4 A.docx
© 2014 Laureate Education, Inc. Page 1 of 5 Week 4 A.docxgerardkortney
 
Assessment 3 ContextYou will review the theory, logic, and a.docx
Assessment 3 ContextYou will review the theory, logic, and a.docxAssessment 3 ContextYou will review the theory, logic, and a.docx
Assessment 3 ContextYou will review the theory, logic, and a.docxgalerussel59292
 
PAGE O&M Statistics – Inferential Statistics Hypothesis Test.docx
PAGE  O&M Statistics – Inferential Statistics Hypothesis Test.docxPAGE  O&M Statistics – Inferential Statistics Hypothesis Test.docx
PAGE O&M Statistics – Inferential Statistics Hypothesis Test.docxgerardkortney
 
Hypothesis Testing Definitions A statistical hypothesi.docx
Hypothesis Testing  Definitions A statistical hypothesi.docxHypothesis Testing  Definitions A statistical hypothesi.docx
Hypothesis Testing Definitions A statistical hypothesi.docxwilcockiris
 
REVISION SLIDES 2.pptx
REVISION SLIDES 2.pptxREVISION SLIDES 2.pptx
REVISION SLIDES 2.pptxssuser0bb1f22
 
BUS 308 Week 3 Lecture 1 Examining Differences - Continued.docx
BUS 308 Week 3 Lecture 1 Examining Differences - Continued.docxBUS 308 Week 3 Lecture 1 Examining Differences - Continued.docx
BUS 308 Week 3 Lecture 1 Examining Differences - Continued.docxcurwenmichaela
 
BUS308 – Week 5 Lecture 1 A Different View Expected Ou.docx
BUS308 – Week 5 Lecture 1 A Different View Expected Ou.docxBUS308 – Week 5 Lecture 1 A Different View Expected Ou.docx
BUS308 – Week 5 Lecture 1 A Different View Expected Ou.docxcurwenmichaela
 
Data analysis powerpoint
Data analysis powerpointData analysis powerpoint
Data analysis powerpointjamiebrandon
 
Commonly used Statistics in Medical Research Handout
Commonly used Statistics in Medical Research HandoutCommonly used Statistics in Medical Research Handout
Commonly used Statistics in Medical Research HandoutPat Barlow
 
Quantitative_analysis.ppt
Quantitative_analysis.pptQuantitative_analysis.ppt
Quantitative_analysis.pptmousaderhem1
 
1. complete stats notes
1. complete stats notes1. complete stats notes
1. complete stats notesBob Smullen
 
Chapter 12Choosing an Appropriate Statistical TestiStockph.docx
Chapter 12Choosing an Appropriate Statistical TestiStockph.docxChapter 12Choosing an Appropriate Statistical TestiStockph.docx
Chapter 12Choosing an Appropriate Statistical TestiStockph.docxmccormicknadine86
 
Inferential AnalysisChapter 20NUR 6812Nursing Research
Inferential AnalysisChapter 20NUR 6812Nursing ResearchInferential AnalysisChapter 20NUR 6812Nursing Research
Inferential AnalysisChapter 20NUR 6812Nursing ResearchLaticiaGrissomzz
 
Inferential AnalysisChapter 20NUR 6812Nursing Research
Inferential AnalysisChapter 20NUR 6812Nursing ResearchInferential AnalysisChapter 20NUR 6812Nursing Research
Inferential AnalysisChapter 20NUR 6812Nursing ResearchLizbethQuinonez813
 
Statistical test
Statistical test Statistical test
Statistical test As Siyam
 
Happiness Data SetAuthor Jackson, S.L. (2017) Statistics plain
Happiness Data SetAuthor Jackson, S.L. (2017) Statistics plain Happiness Data SetAuthor Jackson, S.L. (2017) Statistics plain
Happiness Data SetAuthor Jackson, S.L. (2017) Statistics plain ShainaBoling829
 
Analyzing quantitative data
Analyzing quantitative dataAnalyzing quantitative data
Analyzing quantitative datamostafasharafiye
 

Semelhante a Hypothesis Testing, Effect Size, and Power Analysis (20)

Meta analysis with R
Meta analysis with RMeta analysis with R
Meta analysis with R
 
© 2014 Laureate Education, Inc. Page 1 of 5 Week 4 A.docx
© 2014 Laureate Education, Inc.   Page 1 of 5  Week 4 A.docx© 2014 Laureate Education, Inc.   Page 1 of 5  Week 4 A.docx
© 2014 Laureate Education, Inc. Page 1 of 5 Week 4 A.docx
 
Assessment 3 ContextYou will review the theory, logic, and a.docx
Assessment 3 ContextYou will review the theory, logic, and a.docxAssessment 3 ContextYou will review the theory, logic, and a.docx
Assessment 3 ContextYou will review the theory, logic, and a.docx
 
PAGE O&M Statistics – Inferential Statistics Hypothesis Test.docx
PAGE  O&M Statistics – Inferential Statistics Hypothesis Test.docxPAGE  O&M Statistics – Inferential Statistics Hypothesis Test.docx
PAGE O&M Statistics – Inferential Statistics Hypothesis Test.docx
 
Hypothesis Testing Definitions A statistical hypothesi.docx
Hypothesis Testing  Definitions A statistical hypothesi.docxHypothesis Testing  Definitions A statistical hypothesi.docx
Hypothesis Testing Definitions A statistical hypothesi.docx
 
REVISION SLIDES 2.pptx
REVISION SLIDES 2.pptxREVISION SLIDES 2.pptx
REVISION SLIDES 2.pptx
 
BUS 308 Week 3 Lecture 1 Examining Differences - Continued.docx
BUS 308 Week 3 Lecture 1 Examining Differences - Continued.docxBUS 308 Week 3 Lecture 1 Examining Differences - Continued.docx
BUS 308 Week 3 Lecture 1 Examining Differences - Continued.docx
 
BUS308 – Week 5 Lecture 1 A Different View Expected Ou.docx
BUS308 – Week 5 Lecture 1 A Different View Expected Ou.docxBUS308 – Week 5 Lecture 1 A Different View Expected Ou.docx
BUS308 – Week 5 Lecture 1 A Different View Expected Ou.docx
 
Data analysis powerpoint
Data analysis powerpointData analysis powerpoint
Data analysis powerpoint
 
Commonly used Statistics in Medical Research Handout
Commonly used Statistics in Medical Research HandoutCommonly used Statistics in Medical Research Handout
Commonly used Statistics in Medical Research Handout
 
Quantitative_analysis.ppt
Quantitative_analysis.pptQuantitative_analysis.ppt
Quantitative_analysis.ppt
 
Analyzing Results
Analyzing ResultsAnalyzing Results
Analyzing Results
 
1. complete stats notes
1. complete stats notes1. complete stats notes
1. complete stats notes
 
Chapter 12Choosing an Appropriate Statistical TestiStockph.docx
Chapter 12Choosing an Appropriate Statistical TestiStockph.docxChapter 12Choosing an Appropriate Statistical TestiStockph.docx
Chapter 12Choosing an Appropriate Statistical TestiStockph.docx
 
Inferential AnalysisChapter 20NUR 6812Nursing Research
Inferential AnalysisChapter 20NUR 6812Nursing ResearchInferential AnalysisChapter 20NUR 6812Nursing Research
Inferential AnalysisChapter 20NUR 6812Nursing Research
 
Inferential AnalysisChapter 20NUR 6812Nursing Research
Inferential AnalysisChapter 20NUR 6812Nursing ResearchInferential AnalysisChapter 20NUR 6812Nursing Research
Inferential AnalysisChapter 20NUR 6812Nursing Research
 
Statistical test
Statistical test Statistical test
Statistical test
 
Happiness Data SetAuthor Jackson, S.L. (2017) Statistics plain
Happiness Data SetAuthor Jackson, S.L. (2017) Statistics plain Happiness Data SetAuthor Jackson, S.L. (2017) Statistics plain
Happiness Data SetAuthor Jackson, S.L. (2017) Statistics plain
 
Parametric tests
Parametric  testsParametric  tests
Parametric tests
 
Analyzing quantitative data
Analyzing quantitative dataAnalyzing quantitative data
Analyzing quantitative data
 

Mais de Musfera Nara Vadia

CHARACTERISTICS OF A GOOD INSTRUMENT
CHARACTERISTICS OF A GOOD INSTRUMENTCHARACTERISTICS OF A GOOD INSTRUMENT
CHARACTERISTICS OF A GOOD INSTRUMENTMusfera Nara Vadia
 
Classic and Modern Philosophy: Rationalism and Empicism
Classic and Modern Philosophy: Rationalism and EmpicismClassic and Modern Philosophy: Rationalism and Empicism
Classic and Modern Philosophy: Rationalism and EmpicismMusfera Nara Vadia
 
Single-subject Experimental Research
Single-subject Experimental ResearchSingle-subject Experimental Research
Single-subject Experimental ResearchMusfera Nara Vadia
 
RPP B.Inggris kelas X KTSP Personal Letter ~ writtingskill
RPP B.Inggris kelas X KTSP Personal Letter ~ writtingskillRPP B.Inggris kelas X KTSP Personal Letter ~ writtingskill
RPP B.Inggris kelas X KTSP Personal Letter ~ writtingskillMusfera Nara Vadia
 
RPP B.Inggris kelas X KTSP COMPLIMENT
RPP B.Inggris kelas X KTSP COMPLIMENTRPP B.Inggris kelas X KTSP COMPLIMENT
RPP B.Inggris kelas X KTSP COMPLIMENTMusfera Nara Vadia
 
TELLING ABOUT YOURSELF - LISTENING MATERIAL FOR ICT CLASS
TELLING ABOUT YOURSELF - LISTENING MATERIAL FOR ICT CLASSTELLING ABOUT YOURSELF - LISTENING MATERIAL FOR ICT CLASS
TELLING ABOUT YOURSELF - LISTENING MATERIAL FOR ICT CLASSMusfera Nara Vadia
 
English for Young Learners - Teaching Speaking
English for Young Learners - Teaching SpeakingEnglish for Young Learners - Teaching Speaking
English for Young Learners - Teaching SpeakingMusfera Nara Vadia
 
English for Young Learners - Children Development in Term of Cognitive Develo...
English for Young Learners - Children Development in Term of Cognitive Develo...English for Young Learners - Children Development in Term of Cognitive Develo...
English for Young Learners - Children Development in Term of Cognitive Develo...Musfera Nara Vadia
 
Teaching Oral Communication Skills
Teaching Oral Communication SkillsTeaching Oral Communication Skills
Teaching Oral Communication SkillsMusfera Nara Vadia
 
SEMANTICS AND PRAGMATICS - PRESUPPOSITIONS AND ENTAILMENTS
SEMANTICS AND PRAGMATICS - PRESUPPOSITIONS AND ENTAILMENTSSEMANTICS AND PRAGMATICS - PRESUPPOSITIONS AND ENTAILMENTS
SEMANTICS AND PRAGMATICS - PRESUPPOSITIONS AND ENTAILMENTSMusfera Nara Vadia
 
Technique of Developing Materials
Technique of Developing MaterialsTechnique of Developing Materials
Technique of Developing MaterialsMusfera Nara Vadia
 
Assessing grammar & vocabulary
Assessing grammar & vocabularyAssessing grammar & vocabulary
Assessing grammar & vocabularyMusfera Nara Vadia
 
Bimbingan dan konseling- bidang pengembangan layanan BK
Bimbingan dan konseling- bidang pengembangan layanan BKBimbingan dan konseling- bidang pengembangan layanan BK
Bimbingan dan konseling- bidang pengembangan layanan BKMusfera Nara Vadia
 
ELTMM : Types of instruction material and media and its characteristicsti
ELTMM : Types of instruction material and media and its characteristicstiELTMM : Types of instruction material and media and its characteristicsti
ELTMM : Types of instruction material and media and its characteristicstiMusfera Nara Vadia
 

Mais de Musfera Nara Vadia (20)

CHARACTERISTICS OF A GOOD INSTRUMENT
CHARACTERISTICS OF A GOOD INSTRUMENTCHARACTERISTICS OF A GOOD INSTRUMENT
CHARACTERISTICS OF A GOOD INSTRUMENT
 
Classic and Modern Philosophy: Rationalism and Empicism
Classic and Modern Philosophy: Rationalism and EmpicismClassic and Modern Philosophy: Rationalism and Empicism
Classic and Modern Philosophy: Rationalism and Empicism
 
Single-subject Experimental Research
Single-subject Experimental ResearchSingle-subject Experimental Research
Single-subject Experimental Research
 
RPP B.Inggris kelas X KTSP Personal Letter ~ writtingskill
RPP B.Inggris kelas X KTSP Personal Letter ~ writtingskillRPP B.Inggris kelas X KTSP Personal Letter ~ writtingskill
RPP B.Inggris kelas X KTSP Personal Letter ~ writtingskill
 
RPP B.Inggris kelas X KTSP COMPLIMENT
RPP B.Inggris kelas X KTSP COMPLIMENTRPP B.Inggris kelas X KTSP COMPLIMENT
RPP B.Inggris kelas X KTSP COMPLIMENT
 
TELLING ABOUT YOURSELF - LISTENING MATERIAL FOR ICT CLASS
TELLING ABOUT YOURSELF - LISTENING MATERIAL FOR ICT CLASSTELLING ABOUT YOURSELF - LISTENING MATERIAL FOR ICT CLASS
TELLING ABOUT YOURSELF - LISTENING MATERIAL FOR ICT CLASS
 
How children learn language
How children learn languageHow children learn language
How children learn language
 
English for Young Learners - Teaching Speaking
English for Young Learners - Teaching SpeakingEnglish for Young Learners - Teaching Speaking
English for Young Learners - Teaching Speaking
 
English for Young Learners - Children Development in Term of Cognitive Develo...
English for Young Learners - Children Development in Term of Cognitive Develo...English for Young Learners - Children Development in Term of Cognitive Develo...
English for Young Learners - Children Development in Term of Cognitive Develo...
 
Teaching Oral Communication Skills
Teaching Oral Communication SkillsTeaching Oral Communication Skills
Teaching Oral Communication Skills
 
TEFL- Teaching Vocabulary
TEFL- Teaching VocabularyTEFL- Teaching Vocabulary
TEFL- Teaching Vocabulary
 
SEMANTICS AND PRAGMATICS - PRESUPPOSITIONS AND ENTAILMENTS
SEMANTICS AND PRAGMATICS - PRESUPPOSITIONS AND ENTAILMENTSSEMANTICS AND PRAGMATICS - PRESUPPOSITIONS AND ENTAILMENTS
SEMANTICS AND PRAGMATICS - PRESUPPOSITIONS AND ENTAILMENTS
 
Technique of Developing Materials
Technique of Developing MaterialsTechnique of Developing Materials
Technique of Developing Materials
 
Assessing grammar & vocabulary
Assessing grammar & vocabularyAssessing grammar & vocabulary
Assessing grammar & vocabulary
 
Assessing Writing
Assessing WritingAssessing Writing
Assessing Writing
 
Bimbingan dan konseling- bidang pengembangan layanan BK
Bimbingan dan konseling- bidang pengembangan layanan BKBimbingan dan konseling- bidang pengembangan layanan BK
Bimbingan dan konseling- bidang pengembangan layanan BK
 
Sequence Pattern
Sequence PatternSequence Pattern
Sequence Pattern
 
Speaking to Inform
Speaking to InformSpeaking to Inform
Speaking to Inform
 
Competence-Based Curriculum
Competence-Based CurriculumCompetence-Based Curriculum
Competence-Based Curriculum
 
ELTMM : Types of instruction material and media and its characteristicsti
ELTMM : Types of instruction material and media and its characteristicstiELTMM : Types of instruction material and media and its characteristicsti
ELTMM : Types of instruction material and media and its characteristicsti
 

Último

ROLES IN A STAGE PRODUCTION in arts.pptx
ROLES IN A STAGE PRODUCTION in arts.pptxROLES IN A STAGE PRODUCTION in arts.pptx
ROLES IN A STAGE PRODUCTION in arts.pptxVanesaIglesias10
 
Integumentary System SMP B. Pharm Sem I.ppt
Integumentary System SMP B. Pharm Sem I.pptIntegumentary System SMP B. Pharm Sem I.ppt
Integumentary System SMP B. Pharm Sem I.pptshraddhaparab530
 
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdf
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdfVirtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdf
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdfErwinPantujan2
 
Student Profile Sample - We help schools to connect the data they have, with ...
Student Profile Sample - We help schools to connect the data they have, with ...Student Profile Sample - We help schools to connect the data they have, with ...
Student Profile Sample - We help schools to connect the data they have, with ...Seán Kennedy
 
Concurrency Control in Database Management system
Concurrency Control in Database Management systemConcurrency Control in Database Management system
Concurrency Control in Database Management systemChristalin Nelson
 
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...Nguyen Thanh Tu Collection
 
How to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERPHow to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERPCeline George
 
How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17Celine George
 
What is Model Inheritance in Odoo 17 ERP
What is Model Inheritance in Odoo 17 ERPWhat is Model Inheritance in Odoo 17 ERP
What is Model Inheritance in Odoo 17 ERPCeline George
 
4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptx4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptxmary850239
 
ANG SEKTOR NG agrikultura.pptx QUARTER 4
ANG SEKTOR NG agrikultura.pptx QUARTER 4ANG SEKTOR NG agrikultura.pptx QUARTER 4
ANG SEKTOR NG agrikultura.pptx QUARTER 4MiaBumagat1
 
Choosing the Right CBSE School A Comprehensive Guide for Parents
Choosing the Right CBSE School A Comprehensive Guide for ParentsChoosing the Right CBSE School A Comprehensive Guide for Parents
Choosing the Right CBSE School A Comprehensive Guide for Parentsnavabharathschool99
 
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYKayeClaireEstoconing
 
Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Mark Reed
 
GRADE 4 - SUMMATIVE TEST QUARTER 4 ALL SUBJECTS
GRADE 4 - SUMMATIVE TEST QUARTER 4 ALL SUBJECTSGRADE 4 - SUMMATIVE TEST QUARTER 4 ALL SUBJECTS
GRADE 4 - SUMMATIVE TEST QUARTER 4 ALL SUBJECTSJoshuaGantuangco2
 
Music 9 - 4th quarter - Vocal Music of the Romantic Period.pptx
Music 9 - 4th quarter - Vocal Music of the Romantic Period.pptxMusic 9 - 4th quarter - Vocal Music of the Romantic Period.pptx
Music 9 - 4th quarter - Vocal Music of the Romantic Period.pptxleah joy valeriano
 
Keynote by Prof. Wurzer at Nordex about IP-design
Keynote by Prof. Wurzer at Nordex about IP-designKeynote by Prof. Wurzer at Nordex about IP-design
Keynote by Prof. Wurzer at Nordex about IP-designMIPLM
 
Karra SKD Conference Presentation Revised.pptx
Karra SKD Conference Presentation Revised.pptxKarra SKD Conference Presentation Revised.pptx
Karra SKD Conference Presentation Revised.pptxAshokKarra1
 

Último (20)

ROLES IN A STAGE PRODUCTION in arts.pptx
ROLES IN A STAGE PRODUCTION in arts.pptxROLES IN A STAGE PRODUCTION in arts.pptx
ROLES IN A STAGE PRODUCTION in arts.pptx
 
Integumentary System SMP B. Pharm Sem I.ppt
Integumentary System SMP B. Pharm Sem I.pptIntegumentary System SMP B. Pharm Sem I.ppt
Integumentary System SMP B. Pharm Sem I.ppt
 
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdf
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdfVirtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdf
Virtual-Orientation-on-the-Administration-of-NATG12-NATG6-and-ELLNA.pdf
 
Student Profile Sample - We help schools to connect the data they have, with ...
Student Profile Sample - We help schools to connect the data they have, with ...Student Profile Sample - We help schools to connect the data they have, with ...
Student Profile Sample - We help schools to connect the data they have, with ...
 
Concurrency Control in Database Management system
Concurrency Control in Database Management systemConcurrency Control in Database Management system
Concurrency Control in Database Management system
 
Raw materials used in Herbal Cosmetics.pptx
Raw materials used in Herbal Cosmetics.pptxRaw materials used in Herbal Cosmetics.pptx
Raw materials used in Herbal Cosmetics.pptx
 
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
 
How to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERPHow to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERP
 
How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17
 
What is Model Inheritance in Odoo 17 ERP
What is Model Inheritance in Odoo 17 ERPWhat is Model Inheritance in Odoo 17 ERP
What is Model Inheritance in Odoo 17 ERP
 
4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptx4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptx
 
ANG SEKTOR NG agrikultura.pptx QUARTER 4
ANG SEKTOR NG agrikultura.pptx QUARTER 4ANG SEKTOR NG agrikultura.pptx QUARTER 4
ANG SEKTOR NG agrikultura.pptx QUARTER 4
 
Choosing the Right CBSE School A Comprehensive Guide for Parents
Choosing the Right CBSE School A Comprehensive Guide for ParentsChoosing the Right CBSE School A Comprehensive Guide for Parents
Choosing the Right CBSE School A Comprehensive Guide for Parents
 
YOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptx
YOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptxYOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptx
YOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptx
 
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
 
Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)
 
GRADE 4 - SUMMATIVE TEST QUARTER 4 ALL SUBJECTS
GRADE 4 - SUMMATIVE TEST QUARTER 4 ALL SUBJECTSGRADE 4 - SUMMATIVE TEST QUARTER 4 ALL SUBJECTS
GRADE 4 - SUMMATIVE TEST QUARTER 4 ALL SUBJECTS
 
Music 9 - 4th quarter - Vocal Music of the Romantic Period.pptx
Music 9 - 4th quarter - Vocal Music of the Romantic Period.pptxMusic 9 - 4th quarter - Vocal Music of the Romantic Period.pptx
Music 9 - 4th quarter - Vocal Music of the Romantic Period.pptx
 
Keynote by Prof. Wurzer at Nordex about IP-design
Keynote by Prof. Wurzer at Nordex about IP-designKeynote by Prof. Wurzer at Nordex about IP-design
Keynote by Prof. Wurzer at Nordex about IP-design
 
Karra SKD Conference Presentation Revised.pptx
Karra SKD Conference Presentation Revised.pptxKarra SKD Conference Presentation Revised.pptx
Karra SKD Conference Presentation Revised.pptx
 

Hypothesis Testing, Effect Size, and Power Analysis

  • 1. Chapter 4 Changing the Way We Do: Hypothesis Testing, Power, Effect Size, and Other Misunderstood Issues Annisa Fitri Irwan Della Oferischa Musfera Nara Vadia Rezky Jafri
  • 2. There are a few important steps that researchers can take to make the statistical results from their studies meaningful and useful. They are:  Perform a power analysis before undertaking a study in order to determine the number of participants that should be included in a study to ensure an adequate level of power (power should be higher than .50 and ideally .80). Briefly put, power is the probability that you will find differences between groups or relationships among variables if they actually do exist.  Never set an alpha level lower than .05 and try to set it higher, to .10 if at all acceptable to the research community one is working in.  Report effect sizes and their interpretation.  Report confidence intervals.
  • 3. 1. Null Hypothesis Significant Tests  Null hypothesis (Ho) is there is no difference between groups or that there is no relationship between variables.  Once we reject null hypothesis we will be able to accept alternative hypothesis (Ha).
  • 4. Example:  HR: Will 15 minutes of practice of meaningful drills result in more accurate grammar scores than 15 minutes of telling a story where the grammar in question must be used?  H0 : There is no [statistical] difference between a group which practices grammar using explicit meaningful drills for 15 minutes each day and a group which uses grammar implicitly by telling stories where the grammar is needed for 15 minutes each day.  Ha: There is a [statistical] difference between the explicit and implicit group.
  • 5.  Since only two groups are being compared, a t-test can be used. The t-test statistic is calculated based on three pieces of information.: the mean scores of the groups, their variances, and the size of each group (the sample size).  In NHST process, we should have already decided on a cut-off level that we will use to consider the results of the statistic test extreme. This is called alpha level or significant level.  Baayen (2008) “if the p-value is lower than alpha level we set, we reject the null hypothesis and accept the alternative hypothesis that there is a difference between the two groups (it does not necessarily mean the alternative hypothesis is correct)”  P-value: the probability of finding a [insert statistic name here] this large or larger if the null hypothesis were true is [insert p-value].
  • 6. One-Tailed versus Two-Tailed Tests of Hypothesis  if the only thing we care about is just one of the possibilities, then we can use a one-tailed test. A one-tailed or directional test of a hypothesis looks only at one end of the distribution. A one-tailed test will have more power to find differences, because it can allocate the entire alpha level to one side of the distribution and not have to split it up between both ends.  Two-tailed hypothesis is examining two possibilities of groups. The hypothesis could go in either direction. For example: in our null hypothesis, we would be examining both the possibility that the explicit group was better than the implicit group and the possibility that the explicit group was worse than the implicit group.
  • 7. Outcomes of Statistical Testing True situation in the populationOutcomeobservedinstudy No effect exist No effect exist Effect exist Correct situation (probability= I- α) Type II error (probability=ß) Effect exist Type I error (probability=α) Correct situation (probability= I- ß) Power
  • 8. Type I error (being overeager): concluding there is a relationship when there is none.  Set type I error level by setting alpha (α) level o Commonly set at α = .05 o Possibility of a Type I error is thus 5% Type error II (being overly cautious): concluding there is no relationship when there is one.  Set type II error level (ß) and then calculate power (power = I – ß) o Commonly set at ß = .20, resulting in power =.80 o Possibility of Type II error is thus 20%  Avoid low power by: o Having adequate sample sizes o Using reliable dependent variable o controlling for individual differences o Including pre-test o Using longer post-test o Making sure not to violate statistical assumptions.
  • 9. Problems with NHST There are problems with using NHST method of making conclusions about experiments.  One is that some authors interpret a low p-value as indicate of a very strong result. On the other hand, probably a lower p-value does not make a study more significant, in the generally accepted sense of being important.  On the other hand, the p-value of a study is an index of group size and the power of study, and this is why a p- value of .049 and a p-value of .001 are not equivalent, although both are lower than α=.05
  • 10. Change the Way I Do Statistics  Reporting exact p-values (unless they are so small it would take too much room to report them)  Talking about “statistical” results instead of “significant” or “statistically significant” results  Providing confidence intervals and effect size whenever possible.
  • 11. 2. Power Analysis  Power is the probability of detecting a statistical result when there are in fact differences between groups or relationships between variables.  Power often translates into the probability that the test will lead to a correct conclusion about the null hypothesis.
  • 12. What are the theoretical implications if power is not high?  If the power of a test is .50, this means that there is only a 50% chance that a true effect will be detected. In other words, even though there is in fact an effect for some treatment, the researcher has only a 50/50 chance of finding that effect.
  • 13. What is the optimal level of power?  Power should be above .50 and would be judged adequate at .80. a power level of .80 would mean that four out of five times a real effect in the population will be found.  Power levels ought to be calculated before a study is done, and not after.
  • 14. Help with calculating power using R  How to use the arguments for the “pwr” library, how to calculate effect sizes, and Cohen’s guidelines as to the magnitude of effect sizes. Cohen meant for these guidelines to be a help to those who may not know how to start doing power analyses, but once you have a better idea of effect sizes you may be able to make your own guidelines about what constitutes a small, medium, or large effect size for the particular question you are studying.  Remember that obtaining a small effect size means that you think the difference between groups is going to be quite small.
  • 15. 3. Effect Size  Effect size is the magnitude of the impact of the independent variable on the dependent variable.  An effect size gives the researcher insight into the size of the difference between groups is important or negligible.  If the effect size is quite small, then it may make sense to simply discount the findings as unimportant, even if they are statistical.  If the effect size is large, then the researcher has found something that it is important to understand.  Effect sizes do not change no matter how many participants there are, that makes effect sizes a valuable piece of information, much more valuable than the question of whether a statistical test is “significant” or not.
  • 16. Understanding Effect Size Measures Huberty (2002) divided effect sizes into two broad families: group difference indexes and relationship indexes. Both the group difference and relationship effect sizes are ways to provide a standardized measure of the strength of the effect that is found.
  • 17.  A group difference index, or mean difference measure, has been called the d family of effect sizes by Rosenthal (1994). The prototypical effect size measure in this family is Cohen’s d. Cohen’s d measures the difference between two independent sample means, and expresses how large the difference is in standard deviations.  Relationship indexes, also called the r family of effect sizes, measure how much an independent and dependent variable vary together or, in other words, the amount of covariation in the two variables. The more closely the two variables are related, the higher the effect size.
  • 18. Calculating Effect Size for Power Analysis How to determine the effect size to expect?  The best way to do this is to look at effect sizes from previous in this area to see what the magnitude of effect sizes has been. If there is none, the researcher must make an educated guess of the size of the effect that they will find acceptable, or use Cohen’s effect size.  Cohen notes that effect size are likely to be small when one is undertaking research in an area that has been little studied, as researchers may not know yet what kinds of variables they need to control for.  A small effect size is one that is not visible to the naked eye but exists nevertheless.  Cohen notes that effect size magnitudes depend on the area.
  • 19. Calculating Effect Size Summary  In general, statistical tests which include a categorical variable that divides the sample into groups, such as the t-test or ANOVA, use the d family of the effect sizes to measure effect size. The basic idea of effect sizes in the d family is to look at the difference between the means of two groups, as in µA-µB.
  • 20. Table 4.6 Options for Computing Standardizers (the Denominator for d Family Effect Sizes)  A The standard deviation of one of the groups, perhaps most typically the control group  B The pooled standard deviation of [only the groups] being compared  C The pooled standard deviation [of all the groups] in the design
  • 21. 4. Confidence Intervals  The confidence interval represents “a range of plausible values for the corresponding parameter”, whether that parameter be the true mean, the difference in scores or whatever.  The width of the confidence interval indicates the precision with which the difference can be calculated or, more precisely, the amount of sampling error.  If there is a lot of sampling error in a study, then confidence intervals will be wide, and the statistical results may not be very good estimates.
  • 22. Power through Replication and Belief in the “Law of Small Numbers”  Tversky and Kahneman (1971) point out that replication studies will ideally have a larger number of participants than the original study. With a sample size that is larger than the original, the experimenter will have a better chance of finding a significant result. Sample sizes play a direct role in the amount of power that a study has, and also directly affect the p-value of the test statistic.  Tversky and Kahneman’s proposed law of small numbers states that “the law of large numbers applies to small numbers as well”  In other words, researchers believe that even a small sampling should represent the whole population well, which leads to an unfounded confidence in results found with small sample sizes.
  • 23. Larson-Hall, Jenifer. 2010. A Guide to doing Statistic in Second Language Research Using SPSS. New York: Routledge