Advanced Research Methods Exam Questions and Verified Answers

Advanced Research Methods Exam Questions and Verified Answers a (known as the constant or intercept) - ANSWER-is the point where the line intercepts the y-axis (the value of Y when X is 0) .A dichotomous choice response format is? - ANSWER-a yes/no response, and is the simplest closed-format response - this can be used when there is a clear yes/no answer - it is not sensitive - normally used for personality scales .A factorial design written 3 x 4 is? - ANSWER-a two-way design with three levels of factor A and 4 levels of factor B) - it would have 12 cells .A good scale maximises which observed score? - ANSWER-the true score .a high communality indicates? - ANSWER-that the variable is well-represented by components .A KMO of what is considered satisfactory? - ANSWER- .60 .A likert scale is - ANSWER-a set of items, utilising a numerical rating scale where responses are summed to give a final score .A low communality indicates? - ANSWER-an outlying variable that should be eliminated from the analysis ( .30) .A measure that is reliable but not valid - ANSWER-has a high degree of systematic error, and equal amounts of true score and random error .A mixed design is? - ANSWER-a design in which there is a mixed between-subjects and within-subjects factors .A negative relationship means? - ANSWER-variables covary in opposite directions .A non-relationship means? - ANSWER-variables covary independently .A partial correlation - ANSWER-is the relationship between X1 and Y when shared variance with other X variables has been removed from both X1 and Y (and so on for other X variables) .A perfect relationship means? - ANSWER-one variable predicts another with 100% accuracy .A population - ANSWER-is the entire group you are seeking to describe or understand .A positive relationship means? - ANSWER-that variables covary together .A questionnaire layout - ANSWER-(i) needs to be optimal for a self-administrated questionnaire (ii) needs to appeal to the respondent (iii) needs to not be pressing when a questionnaire is administered through an interviewer .A reliable and valid scale has a - ANSWER-high true score, low systematic error and low random error .A sample - ANSWER-is a subset of the population, from which you want to infer population characteristics .A semi-partial (or part) correlation - ANSWER-is the relationship between X1, and Y when shared variance with other X variables has been removed from X1 (and so on for other X variables) .A simple effect of a factor is - ANSWER-its effect at only one level of the other factor in a factorial design. - it is a one way ANOVA .A useful indicator of effect size is? - ANSWER-the odds ratio, which can be calculated as number in first group/number in second group (variable) .A x (B x C x S) is? - ANSWER-three factors, one between groups variable (A); two repeated (B and C) .A x B x (C x D x S) is? - ANSWER-4 factors, two between groups (A and B); two repeated (C and D) .A x B x (C x S) is? - ANSWER-3 factors, 2 between-groups (A, B); one repeated C .adjusted standardised residuals are - ANSWER-for each cell, the observed minus expected frequency divided by an estimate of the standard error .Administration methods have implications for? - ANSWER-the representativeness of the sample and the extent to which the survey will be able to provide accurate estimates of populations parameters. .an A x B design is? - ANSWER-the simplest form of two-way factorial ANOVA .an A x B x C design is? - ANSWER-a three-way, with 3 IVs factorial ANOVA .An ANOVA can be used with surveys when - ANSWER-(i) surveys contain an experimental manipulation (ii) to assess differences as a function of demographic variables, treated as the IV (these are natural-groups correlational designs, not experimental designs, and do not allow causal inference) .An eigenvalue represents? - ANSWER-the proportion of variance in the set of variables accounted for by a component or factor .An interviewer must have neutrality and - ANSWER-(i) must remain neutral throughout interview (ii) be unbiased in instructions and clarification (iii) convey no hints of one's own opinion .An ordinal interaction is - ANSWER-an interaction in which the lines on a line graph are not parallel, but they also do not cross over - in these interactions, main effects are still important they just do not tell the whole story .An unreliable scale has high - ANSWER-random error, a low true score, and a low systematic error .Anonymity - ANSWER-(i) improves data quality with sensitive topics (ii) is advisable when information is required about age, income level etc. (iii) possible with self-administration but not with interviewer administration .Assumptions of a factorial between-groups ANOVA - ANSWER-1. the DV should be at the interval level or ratio level of measurement 2. the scores are randomly sampled from the populations of interest 3. the score are independent 4. the scores on the DV within each cell are normally distributed in the population 5. the is homogeneity of variance .Assumptions of bivariate correlations - ANSWER-(i) the relationship must be linear (ii) the relationship should homoscedastic (iii) there should be no restricted range (iv) there should be no outliers (v) no extreme groups (vi) randomly sampled .Attitude choice response formats are - ANSWER-a variation of the yes/no forced choice response, and allows the participant to provide a number of options under the yes/no statements - it allows participants to feel listened to and can avoid agree-disagree acquiescence response set bias .b (known as the b-weight or regression coefficient) - ANSWER-is the slope of the line (the amount by which Y increases for every 1 unit increase in X) .By convention, correlations are regarded as follows - ANSWER-.00 - .09 = very weak, negligible .10 - .29 = weak, low, small .30 - .49 = moderate, medium .50 - .69 = strong, high, large .70 - 1.0 = very strong, very high .Can convenience samples be used for hypothesis testing? - ANSWER-Yes: (i) it is assumed that the sample is "like" people in general (ii) assumption may not be tenable (iii) results of hypothesis testing can only be generalised to the population of people who really are "like" those in convenience sampling .chi-square formula - ANSWER- .Chi-square involves - ANSWER-comparing the expected frequencies under the null hypothesis with the actual observed frequencies .Coding missing data - ANSWER-the data may have different missing values to code for different reasons: (i) they didnt understand the question (ii) they chose not to respond .Computer data entry - ANSWER-(i) every participant needs an unique ID for each participant (ii) data entry is easier in excel (iii) fastest in word processing program, some has some issues (iv) always enter data in most original form (v) always save original data in separate file (vi) always keep a project workbook to be used as a codebook .Confirmatory factor analysis is most often used in? - ANSWER-more specialised tests, and is used to test prior theories of the structure of a particular domain, that is, to test (confirm) whether a particular construct does indeed produce the factors it should according to theory .Convenience samples - ANSWER-(i) are obtained by whatever method available (ii) may be biased and unrepresentative of the population (iii) cannot be used to infer population characteristics .Correlation - ANSWER-describes the relationship between two variables; consequently its precise name is bivariate correlation .Correlation can be used for? - ANSWER-inferential statistics that tests hypotheses about the correlation in the population .Correlation coefficients are - ANSWER-statistical measures of correlation .Correlation relationships are best shown by? - ANSWER-a scatterplot or scattergram .Correlations among IVs .90 - ANSWER-indicate multicollinearity, if this is the case, they need to be deleted .Cronbach's alpha automatically becomes larger how? - ANSWER-with the more items there are in an instrument .Cronbachs alpha - ANSWER-corresponds approximately to the "mean of all the split-half coefficients resulting from different splittings of a test" .degrees of freedom for two-way chi-squared - ANSWER-(r-1) (c-1) - where r = number of rows - where c = number of columns .Depth interviews are used to - ANSWER-(i) gain an in-depth understanding of an area (ii) gain a more open format (iii) they are ideally recorded on tape (iv) they are subjected to qualitative analysis (v) the main role is in preliminary exploratory research .Does Pearson's product-moment correlation (r) have the assumption of normality? - ANSWER-yes, it measures whether the variables are normally distributed in the population .Errors in predictions are known as? - ANSWER-residuals, and are the difference between actual Y scores and predicted Y scores (Ŷ) .Every member of the population has - ANSWER-an equal (or known) probability or chance of being selected .expected frequencies formula - ANSWER- .Experimental designs - ANSWER-involve manipulation and control of factors to allow causal conclusions