# Research Inferential Stats

 Inferential statistics are based on the laws of ____________ and are used to estimate population ________ from sample _________ ProbabilityParametersStatistics Given data from a sample, inferential stats allows us to _______________ Draw conclusions or make inferences about a population Different researchers applying inferential statistics to the same data are likely to draw __________ conclusions. the same Inferentials stats assume _______ sampling from populations, assumption that is widely violated. Random ________ is the tendency for statistics to fluctuate from one sample to another. Sampling error _________ is a theoretical distribution of a statistic, using the values of the statistic (the means) computed from an infinite # of samples as the data points in the distribution. Sampling Distribution What is a sampling distribution theoretical? It is not actual because in practice no one draws consecutive samples from a population and plots their means. When a distribution is normal _____ % of values fall between +, - 1 SD from the mean, _____ % fall between =, - 2 SD and _______ fall between =, - 3 SD's.. 68%95%99.7% The standard deviation of a sampling distribution of the mean is called ________ standard error of the mean (SEM) The___________ is an estimate of how much sampling error there is from one sample mean to another. SEM If we increase our sample size, we ________ the accuracy of our estimates. increase Statistical inference consists of two techniques: ________ and ____________. estimation of parametershypothesis testing ____________ is used to estimate a parameter Parameter estimation Examples of parameters are: a meana proportiona mean difference btw groups Estimation can take two forms: 1. point estimation2. interval estimation __________ involves calculating a single descriptive statistic to estimate the population parameter. Point estimation __________ estimation is useful because it indicated a range of values within which the parameter has a specified probability of lying. Interval The ___________ is the range of values within which a population parameter is estimated to lie, at a specified probability (e.g., 95 %__) Confidence Interval (CI) Confidence intervals reflect the researchers' _____________ risk of being wrong. With a 95% CI, researchres aceept the probability that they will be wrong ____ times out of 100. A 99% CI, sets the risk of being wrong at _____. 5. (5%)1%. For proportions based on dichotomous variables (positive or negative for a disease), the theoretical distribution is ___________ Binomial distribution ________ states that there is no relationship among the variables; that any findings are due to chance. Null Hypothesis A _________ error means that the Null Hypothesis was rejected, when in fact it was true. *"False positive".  We said there was a difference/effect when in fact, there was not.* Type 1 A ______ error is made when researchers accept the null hypothesis when  it fact it was NOT true. This is a "false negative" conclusion.  We said there is no difference, when in fact one exists. Type 2. A ________ error might prevent a good drug from coming to market. A_________ error might allow an ineffective drug to come onto the market, while a Type 2Type 1 How do researchers control the risk of a Type 1 error? By selecting a level of significance, which signifies the probability of incorrectly rejecting a true null hypothesis. What is alpha? Level of significance. Usually .01 or .05. With a .05 alpha level, what is the risk of us incorrectly rejecting the null hypothesis? 5 out of 100 chance. Lowering the risk of a Type 1 error can increase___________ the risk of making a Type 2 error. The simplest way of reducing the risk of a Type 2 error is to ___________. Increase sample size Levels of significance are analogous to _____. Whereby alpha of .05 =_______ CI's.  95% CI. What does the word significant mean in statistics? Obtained results are not likely to be due to chance, at a specified level of probability. What does a non-significant result mean? An observed result could reflect chance fluctuations. A one-tailed test would be best when a ________ hypothesis is strongly suspected. Directional There are two broad classes of statistical tests. Parametric and Nonparametric ________ tests involve estimation of a parameter, require measurements on at least an interval scale and involve several assumptions, such as the assumption that the variables are normally distributed in the population. Parametric __________ tests, do not estimate parameters; they involve less restrictive assumptions about the shape of the variables' distribution than do __________ Non-parametric Parametric When the N is> 50, it may not be necessary to use ___________ statistics, unless the population has a markedly unusual distribution. Nonparametric What is the central limit theorem? A statistical principle stipulating that the larger the sample, the more closely the sampling distribution of the mean will approximate a normal distribution and the mean of a sampling distribution equals the population mean. When comparisons involve different people (men versus women), the study uses a _________ design and the statistical test is a ____________. Between-subjects designTest for Independent groups __________ statistical tests are used when there is only one group, that is used in multiple conditions (cross-over designs). Tests for dependent groups What are the steps for testing a hypothesis? 1. Select the appropriate test2. Establish the level of significance3. Select a 1-tailed or 2-tailed test.4. Compute a test statistic5. Determine the degrees of freedom6. Compare the test stat with a tabled value. ________ refers to the # of observations free to vary about a parameter. Degrees of Freedom. What is the parameter procedure for testing differences in group means? t-test What is an adjustment made to establish a more conservative alpha level when multiple statistical tests are being run from the same data set. Bonferroni correction When means for two sets of scores are not independent (dependent), researchers should use________ a paired t-test VandeVusse et al. used ______ to assess changes in women's HR, RR and tensoin anxiety following exposure to a 30 min self-hypnosis intervention. paired t-test In certain 2 group situations, a non-parametric test may be needed. Two examples/reasons to use a non-parametric test are_____,_______ 1. if the dependent variable is on an ordinal scale2. If the distribution is markedly non-normal. The _________ test is the non-parametric analog of an independent group's t-test and involves assigning ranks to the two groups of scores. The sum of the ranks for the 2 groups can be compared by calculating the ____ statistic. Mann-Whitney U TestU When ordinal level data are paired (dependent), the ________ test can be used. This test involves taking the difference between paired scores and ranking the absolute difference. Wilcoxon signed-rank. _______ is the parametric procedure for testing differences between means when there are 3 or more groups by comparing variability between groups to variability within groups. ANOVA The statistic computed in ANOVA tests is _______ F-ratio. Authorepm49 ID311210 Card SetResearch Inferential Stats DescriptionInferential Statisics Updated2015-11-11T00:10:29Z Show Answers