╨╧рб▒с>■  &(■   %                                                                                                                                                                                                                                                                                                                                                                                                                                                ье┴3 Ё┐jbjb^^ Ъh<h<┘ >      lЄЄЄЄЄЄЄ66668n z 6▀ ъТТТТТТТТЬ Ю Ю Ю Ю Ю Ю ,╔ щ Ш╩ ЄТТТТТ╩ VЄЄТТТVVVТ╛ЄТЄТЬ VЄЄЄЄТЬ V2VИ ЄЄИ Ж ╓э╛╝66PИ И ▀ ▀ И Б VБ И V├хTests of statistical significance are used to make an inference about the characteristics of a population from sample data. If the sample is a representative sample from a population, the inference can be applied to that population. If, as is usually the situation in behavioral science research, the sample is a convenience sample, the inference is the to unknown and unspecified population of which the sample is a representative sample. Researchers typically have expectations about the outcome of a study. They expect a particular pattern of results. However, all tests of statistical significance begin with a null hypothesis (Ho) which is always specific and is (almost) always a hypothesis that people will not differ, that there is not correlation or association between the variables, etc. Examples of null hypotheses for a number of statistical tests are shown below. If two Hos are given, one is for a subject variable study and the other for an experiment. One variable chi-square: The number of people preferring each brand of cola will be the same. df = k - 1 Contingency table chi-square: There is no association between gender and living on vs. off campus. df = (r Ц 1)(c Ц 1) Correlation: The true value of the correlation between study time and GPA among college students is 0.00. df = N - 2 Single sample z-test: (Use when pop. st. dev.( () is known.) Austinites do not differ from other people in their mean IQs, ( = 100, (( = 16). Single sample t-test: (Sample data used to estimate st. dev. (S).) People who exercise regularly have the same mean heart beat rate (pulse) as people in general -- 72 beats per minute. df = N - 1 Two sample t-test: (independent samples): There is no difference in the mean walking speed of males and females. Recall is the same when people are given shallow processing as opposed to deep processing instructions. df = N1 Ц 1 + N2 - 1 Two sample t-test (paired scores or repeated measures or within subject designs). Also called direct difference t-test when difference scores are analyzed. Mean SAT scores will not change if students take the test one week and then take another form of the test a second week. PeopleТs reaction time to light is the same as to sound. df = N - 1 One-way analysis of variance (One-way F-test : There is no difference in locus of control scores of freshman, sophomores, juniors, and seniors. Performance is not affected by instructions designed to make the participants low, moderate, or high in anxiety. BGdf = KЦ1, WGdf = K(N-1) where N=number of subjects in each group. Two-way analysis of variance (Two-way F-test) : There is no main effect for variable A; there is no main effect for variable B; there is no interaction of variables A and B. INFERENTIAL STATISTICS --Tests of Statistical Significance 6?l{a b о п ╕ ╣ o p x y ┌¤¤∙ї∙єєЁCJH* jmЁ jsЁ>*║╗╨ ╥ = > ╕ ╣ 1 2 ├ ─ М Н ~  ▀рф)*┘¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤√¤¤¤║╗╨ ╥ = > ╕ ╣ 1 2 ├ ─ М Н ~  ▀рф)*┘¤░╨/ ░р=!░"░#Ра$Ра%░|HHЇ@ ю юRG(№HH╪(d'`Р i(@ё (NormalCJmH <A@Є б<Default Paragraph Font,@Є,Header  ╞р└!, ,Footer  ╞р└!    ш<   X┘<─ ╚=====@   |24дж&(WY_pz∙√АВik╙╒ф ш Ё Ї  ┘   ╥24=>дж╕&(12^apАВМНik~Щ╙╒▀р  q ф ў ∙ ) * V Y ╪ ┘     Clarke A. Burnham<Macintosh HD:Temporary Items:AutoRecovery save of InferentiaClarke A. Burnham;Macintosh HD:418 -- Spring, 98:Inferential Stat -- Overview Psychology<Macintosh HD:Temporary Items:AutoRecovery save of Inferentia PsychologyLMacintosh HD:Desktop Folder:418 -- Spring, 2002:Inferential Stat -- Overview PsychologyLMacintosh HD:Desktop Folder:418 -- Spring, 2002:Inferential Stat -- OverviewClarke Burnham5HomePage:Faculty:Burnham:Inferential Stat -- OverviewClarke Burnham6HomePage:Faculty:Burnham:Inferential Stat -- Overv.htmClarke BurnhamLMacintosh HD:Desktop Folder:418 - Spring, 2004:Inferential Stat -- Overv.htmClarke BurnhamHMacintosh HD:Desktop Folder:418 - Spring, 2004:Inferential Stat -- OvervClarke Burnham2HomePage:Faculty:Burnham:Inferential Stat -- Overv╪   @А─!▓5k @ @GРTimes New Roman5РАSymbol3Р Arial3РTimes"qИЁ╨'hП+ЕfП+Еf│тdСёГ%ЁДе└┤┤А>p√ :B#ЁД▀  INFERENTIAL STATISTICSUniversity of TexasClarke Burnham■  рЕЯЄ∙OhлС+'│┘0МРШ╕─рь№   < H T `lt|Д'INFERENTIAL STATISTICS9NFEUniversity of TexasnivNormaliClarke Burnhame2arMicrosoft Word 9.0@@J"~╠ю┴@ТЩB╒2─@ТЩB╒2─Сё■  ╒═╒Ь.УЧ+,∙о0, hpи░╕└ ╚╨╪р ш '.Department of Psychology, University of TexasIA√ ╧ INFERENTIAL STATISTICS Title ■   ■   ■    !"#$■   ¤   '■   ■   ■                                                                                                                                                                                                                                                                                                                                                           Root Entry         └FЗnл2─)А1Table             WordDocument        ЪSummaryInformation(    DocumentSummaryInformation8            CompObj    XObjectPool            Зnл2─Зnл2─            ■                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                           ■      └FMicrosoft Word Document■   NB6WWord.Document.8