Also keepinmind that whether the signofthe calculatedidvalue is positiveornegativewilldepend onwhich data set you subtract from which. and whether ornotitmatterswilldepend on whether you aretestingaone- or two-tailed alternative hypothesis. We would expect leadership ability to increase (whichimplies directional testing).sothe mean leadership ability after the workshop should be greaterthanmeanleadership abilitybeforethe workshopifthe alternative hypothesis is valid.Question 20The researcher finds that the test statistic (calculatedinQuestion9) yields a p-value ofp = 0.000572 (as calculated by a computer). What conclusion can be drawn from this? (Use a level ofsignificance of1%).1.There is a significant improvementinthe level of assertiveness as measured before and after thetraining2.There is no significant improvementinthe level of asseniveness as measured before and after thetraining3.The probability value is too low, so thenullhypothesis cannot be rejectedwithconfidence4.There is a significant differenceinthe level of assertiveness as measured before and after thetraining.infavour of the former•fiAnswer:Option1iscorrem.The p-value is smaller than the levelofsignificance(thatis. p-value 0.000572<0.01). The small p-valueimplies that there is a very small probability that the difference between the two means (beforeandafter theworkshop) is the result of chanceorrandom measurement error, so the null hypothesis (which wouldimplyno change) can be rejected.Option 2 is incorrectandOption 3 implies that a low p-value is associatedwithnor rejecting thenullhypothesis, which is the wrong way round. The smaller the p-value. the less the chance is that the nullhypothesisisvalid. The answer giveninOption4isinthe wrong direction:itimplies assertiveness isgreater before than aftertheassertiveness training.

48 IP a g eQuestion 21The researcher is also interested to know whetherthedifference between the two assertiveness scoresbefore and afler the workshop isofpract/ca/ significance:thatis to say, irrespectiveofits statisticalsignificance, whether the effect can be regarded asbigorsmall. Calculate the relevant measurement of thesize of the effect to answer the question below.The size of the effect was found to be - - - - -.small2.medium3.large4.impossible to determine from this dataM Answer:Option3is correm.Toget an indicationofthe practical usefulness of the result the elect size should be calculated. This wouldgiveanindicationofthe absolute size of the statistical effect, independentofthe sizesofthe samplesthatwere used (see Section 3.3.3inthe PYC3704 Study guide).Inthe case where two means are beingcompared, this effect size can be foundbycalculating Cohen sd.The general form of this measurementisto subtract one mean from the other and thentodividethisby the overall standard deviation forallthe data(seep.87inthe PYC3704 Study Guide). The actualmayinwhich Cohen's d is to be calculated canhowever differdependingon which mean is to be compared with which (i.e., it differs s lightly depending onwhich statistical test was used for this data).

Upload your study docs or become a

Course Hero member to access this document

Upload your study docs or become a

Course Hero member to access this document

End of preview. Want to read all 203 pages?

Upload your study docs or become a

Course Hero member to access this document