9 views

Uploaded by Daniel Randolph

t

- Natural Resources Damage Assessment - BP Oil Disaster - Gulf of Mexico
- Text book answers unit 11.doc
- Homework-pval
- 02 Tests
- New Microsoft Office Word jnkjv njxzklncvcklj nddlkncx klnkljb nlj
- new-crib2
- Attributes Enhancing Industrial Productivity an Empirical Analysis
- Z test
- Hypothesis testing
- Gross Motor Skills of Children With Mild Intellectual Disabilities
- Credit Rating MBA
- Mba 206
- p_hacking
- Stock_Watson_3U_ExerciseSolutions_Chapter5_Students.pdf
- 9. Prosiding Four a Hasni Flypaper Effect
- Test 1 Prep
- 09_HW_Assignment_Biostats.docx
- Independent Samples t Test (LECTURE).docx
- Test statistics
- Statistical Tests

You are on page 1of 21

The Setting: Individuals in some population can be classified into one of two categories. You want to make inference about the proportion in each category, so you draw a sample. Examples: Drug X is administered to 100 patients with a particular disease. 50 improve. Test whether this drug is better than drug Y, which is known to produce improvement in 45% of patients. In a poll of 500 voters, 200 say that they will support a particular candidate in the election. Give a 95% confidence interval for the proportion of all voters who will support the candidate.

Inferential Procedures: Hypothesis Test - critical value method - p-value method - confidence interval method Confidence Interval

Hypothesis Test: Specify the null and alternative Define rejection region for the test statistic z

H0 : p = p0

HA : p < p0

HA : p p0

HA : p > p0

Test Statistic:

z=

p0 p p 0(1 p 0) n

If H0 is true, z comes from standard normal distribution Critical value method: Choose z* from normal table to define rejection region If z falls in rejection region, reject H0 P-value method: Use the normal table to determine probability of a value at least as extreme as z coming from standard normal distribution. Report this p-value and your conclusion.

Level-C Confidence Interval:

z* p= p

(1 p ) p n

To perform a two-sided test of H0: p = p0 at significance level , you can calculate the level-(1- ) confidence interval and check whether p0 falls within the interval.

Types of two-sample tests: two-sample z-test two-sample t-test two-sample test of proportions

Two-sample z-test

The Setting: You want to determine whether two different populations have the same mean for some variable of interest, so you draw two independent samples. Furthermore, you know the variances in these two populations without error. In real life, this would only tend to happen if your samples were so large (e.g., 5,000) that you could pretend that the sample variances were really the true variances Example: SAT scores for students who use a particular prep course have an SD of 100, while those for students who do not use the course have an SD of 150. You sample 50 students who take the course and 50 who do not. Test whether taking the course changes the mean score for students

Two-sample z-test

Hypothesis Test

HA : x < y

H0 : x = y HA : x y

HA : x > y

z=

X Y

2 x

nx

2 y

ny

Two-sample t-test

The Setting: You want to determine whether two different populations have the same mean value for some variable of interest. You draw two independent samples to test this. Note that in general we will use this test instead of the ztest, since we will not know the population standard deviations without error. Example: You want to determine whether students who take a given prep course have, on average, the same SAT score as those who do not. You sample 50 students who take the course and 50 who do not and record their SAT scores

Two-sample t-test

Hypothesis Test Null and alternative hypotheses same as for two-sample z-test Test statistic is now

t=

X Y s + nx n y

2 x 2 sy

Two-sample t-test

Common Mistake: Sometimes you will have what appear to be two independent samples, but are really two observations each on a single set of units. Use one-sample methods for inference on such data. This sample will consist of the differences between the two values for each subject.

Example: You want to determine whether a drug lowers patient cholesterol. You record the cholesterol levels of 100 volunteers, then administer the drug for three months. After this period you again measure the patients cholesterol values Strategy: Define D = change in a patients cholesterol after taking drug. Now do a one-sample t-test:

H0 : D = 0 HA : D < 0

t=

D0

2 sD nD

The Setting: In two independent populations, individuals can be classified into one of two categories. You draw samples from both populations to make inference about whether the category proportions are the same in both populations Example: You take a sample of 100 government concentrators and 100 concentrators from other fields. Test whether being a government concentrator makes a college student more or less likely to vote.

Hypothesis test:

HA : px < py

H0 : px = py HA : px py

HA : px > py

z=

x p y p

Where

= p

x + ny p y nx p nx + n y

1 1 (1 p ) + p nx n y

Level-C Confidence Interval:

y (1 p y) x (1 p x) p p x p y) z* ( px p y ) = ( p + nx ny

Common Mistake: Notice that in this case the standard errors of the hypothesis test and confidence interval are different!

Chi-square test

The Setting: In two (or more) independent populations, individuals can be classified into two (or more) different categories. You draw samples from all populations to test whether the category an individual falls into is independent of that individuals population. Example: You choose a sample of 100 men and 100 women and ask subject whether s/he watches professional football. Test whether gender is independent of football viewing

Chi-square test

Hypothesis test (Continuing our example): H0: gender is independent of watching football HA: gender is not independent of watching football Remember that we cannot test a one-sided alternative using the chi-square test!

Chi-square test

Procedure: 1. Create the 2x2 table of observed counts 2. Create the 2x2 table of expected counts 3. Calculate the 2 statistic 4. Draw conclusion

Chi-square test

1. Create the 2x2 table of observed counts 2. Create the 2x2 table of expected counts Expected count in a cell =

rc n

Where r = row total for that cell, c = column total for that cell, n = total for whole table

Chi-square test

3. Calculate the 2 statistic

2

Where the summation is over all cells.

Chi-square test

4. Draw conclusion

Under H0, 2 follows a chi-square distribution with (r-1)(c-1) degrees of freedom

To use the critical value method, determine the value x* that cuts off of the tail area. Reject H0 if 2 > x*. To use the p-value method, determine the area to the right of 2 on the chi-square curve. Report this p-value and draw your conclusion.

Area =

Area = p-value

x*

- Natural Resources Damage Assessment - BP Oil Disaster - Gulf of MexicoUploaded byMaureen Dauphinee
- Text book answers unit 11.docUploaded bysmartie_pants
- Homework-pvalUploaded bydavparan
- 02 TestsUploaded byValentin Adam
- New Microsoft Office Word jnkjv njxzklncvcklj nddlkncx klnkljb nljUploaded byBhagwati Shukla
- new-crib2Uploaded bySkander El Amri
- Attributes Enhancing Industrial Productivity an Empirical AnalysisUploaded byInternational Journal of Research in Engineering and Technology
- Z testUploaded byBitoy Aguila
- Hypothesis testingUploaded byMohammad Shaniaz Islam
- Gross Motor Skills of Children With Mild Intellectual DisabilitiesUploaded bynurul_sharaswati
- Credit Rating MBAUploaded byultodangalocal
- Mba 206Uploaded byNaga Mani Merugu
- p_hackingUploaded bykerredai
- Stock_Watson_3U_ExerciseSolutions_Chapter5_Students.pdfUploaded bySubha Mukherjee
- 9. Prosiding Four a Hasni Flypaper EffectUploaded byRisda Yanti
- Test 1 PrepUploaded bySharif M Mizanur Rahman
- 09_HW_Assignment_Biostats.docxUploaded byjhon
- Independent Samples t Test (LECTURE).docxUploaded byMax Santos
- Test statisticsUploaded byTyson Khagai
- Statistical TestsUploaded byEnriqueMinhaz
- 29. Format. Hum- Research Study on Factors Affecting Purchase of Gold, Silver or Diamond Jewellery OnlineUploaded byImpact Journals
- project part 7--reflection and eportfolio- allUploaded byapi-316479494
- sw 4810 fall 2014 august 2014 finalUploaded byapi-246987386
- Impact of Income Components in Determine the Market Value: Evidence from Pakistan Cement IndustryUploaded byFarRukh Ijaz
- Analysis Using Spss 21Uploaded byrrutayisire
- Topic 6 Two Variable Regression Analysis Interval Estimation and Hypothesis TestingUploaded byOliver Tate-Duncan
- statistikUploaded byKelas B Kimia
- 5414ijmvsc02Uploaded byWilliam Scott
- skittles projectUploaded byapi-319345408
- Block fiUploaded byAppTest PI

- State of the Housing Sector by Shri Lalit Kumar, NHBUploaded byDaniel Randolph
- Tertiary (Service) Sector of the EconomyUploaded byDaniel Randolph
- Syllabus BBA 301Uploaded byDaniel Randolph
- Cost Accounting Techniques Adopted by Manufacturing and Service IndustryUploaded byDaniel Randolph
- Anova ProbsUploaded byDaniel Randolph
- Graphical Method in Game TheoryUploaded bymiss_bnm
- 802_32cristodoulou.pdfUploaded byGizachew Zeleke
- 1394614050 Borrower Rating Related .FullUploaded byDaniel Randolph
- 214chpt11.pdfUploaded bySajith Ranatunga
- Claim Both HRA and Loan DeductionUploaded byDaniel Randolph
- Technical Analysis on Equity SharesUploaded byDaniel Randolph
- SIPHandbook 2009 FinalUploaded byDaniel Randolph
- Sarfaesi PapersUploaded byDaniel Randolph
- Credit AppraisalUploaded byDaniel Randolph
- Info About IOBUploaded byDaniel Randolph
- Organizations TypesUploaded byDaniel Randolph
- presentationonleadership-110924023828-phpapp01Uploaded byDaniel Randolph
- Banking Visin 2010Uploaded byaditya086
- Waiting Lines (Class Problems).pdfUploaded bydreanichole
- ERP Overview1Uploaded byNirmal
- Corruption in India 2010 and BeforeUploaded byHaindava Keralam
- Acetone 3Uploaded byNuNu Taofik Nurdiansyach

- Sample size estimation and power analysis for clinical research studies.pdfUploaded bydrnareshchauhan
- FDI and Economic GrowthUploaded byRabiadz Souf
- somers DUploaded byjaxmsb
- Chapter 4 - Thesis PUP Sta. Maria BulacanUploaded byAkira
- CassavaUploaded byDipo Orims
- Hydrolysis of Sorghum Starch for Ethanol ProductionUploaded byNajiah Nadir
- Performance Factor Impact on Profitability of Cementmanufacturing Sector in PakistanUploaded byMuhammed Amjad Islam
- Growth Implications of Managerial Finance in Business: Empirical Evidence from Akwa Ibom State, NigeriaUploaded byPremier Publishers
- t-testUploaded byReycelle Tantoco
- Applying Organizational Concepts in the Iran Public SectorAUploaded bypiklis
- Thesis referenceUploaded byMielle Grace Carballo
- Applied Statistics MCQUploaded byPrashant Sunagar
- STA404 CHAPTER 08Uploaded byIbnu Iyar
- [000298]Uploaded byFakeAccount1
- Fundamental Uncertainity and Stock MarketUploaded byFaisal Khalil
- Ats ProjectUploaded byAnymRai
- Vcd TutorialUploaded byjuntujuntu
- IJEAS0211013Uploaded byerpublication
- A Study of General Mental Ability among Players And Non-Players Of Chandigarh SchoolsUploaded byInternational Journal for Scientific Research and Development - IJSRD
- Scope InsensitivityUploaded byinvisiblhat
- Anova/Post Hoc TestUploaded bycj1891
- Normality Tests in SASUploaded bydkanand86
- Testing of Hypothesis_one SampleUploaded byAnkur
- The Impact of Economic Sanctions on GrowthUploaded bychelia75
- Nakagawa 2007Uploaded byসাদাকালো
- Excel Practical for StatUploaded byedduardoc
- Ahss Guidelines v 23Uploaded byRadesh Vangipuram
- Corporate Social Responsibility and Organization Performance Evidence from JordanUploaded byIJRASETPublications
- MODULE 15 Hypothesis TestingUploaded byEza Hasnan
- PartyUploaded byladzzz123