BSNS112 Exam Questions And Correct Answers, Complete Solution

BSNS112 Exam Questions And Correct Answers, Complete Solution Categorical (Qualitative) data is either: Nominal (stand alone/label); or Ordinal (ordered/ranked) Numerical (Quantitative) data is either: Discrete (whole numbers) Continuous (infinite options) Ways to present categorical data: Bar graphs (percentages labelled), Pie graphs, Cross tabulations, Stacked bar graph, clustered bar charts Range Maximum - minimum Variance Squared deviation around the mean Coefficient of variation Relative measure of variability A skewness value more than _____ its standard error is taken to indicate a ___________ departure from symmetry. twice, significant Left skew Negative skew Right skew Positive skew Kurtosis Extent to which observations cluster around a central point and fatness of tails. Empirical rule Values for Z and t Which distribution is the odd one out w.r.t. skewness? A. Incomes B. Household debt C. Age when people retire D. Capital value of publicly listed companies very strong correlation r 0.9 strong correlation 0.75 r 0.9 moderate correlation 0.4 r 0.75 weak correlation r 0.4 critical value for negative, one-tailed test "" (1 proportion) -1.645 critical value for positive, one-tailed test "" (1 proportion) 1.645 critical value for two tailed test "not equal to" (1 proportion) +/-1.96 unequal variances value of levene's test is less than 0.1 chi-squared test of independence ei=(column total x row total)/total sample size d.f.=(c-1)(r-1) chi-squared goodness of fit ei=n x pi d.f.=k(no. of categories)-1 if test stat is greater than the critical value w/ degrees of freedom reject H0 if the test statistic increases, the p-value will decrease A priori classical probability based on existing information Empirical (relative frequency) work out probability based on experiments Subjective probability that is opinionated standard error standard deviation of the sample mean/proportion with respect to repeated sampling marginal probability total probability of a row or column covariance the measure of the co-movement between two variables non-parametric tests are more likely to produce a type _ error 2 if the confidence interval ________ 0, it is not statistically __________. includes, significant if data is not approximately normally distributed then we must use a ___-__________ test non-parametric Pearson's chi square value means Chi-squared Independence Test F test statistic consists of: variation between samples (location)/variation within samples (shape) The f test statistic is always ________ and ___-______. positive, one-tailed population treatment Non-parametric ANOVA Kruskal-Wallis Test ANOVA Test for Medians Independent Samples Median Test Multiple pairwise tests have a higher likelihood of producing a type _ error 1 Model Framework for looking at reality Concept General idea Method Research technique Methodology General approach to studying research questions Triangulation Using more than one method to collect data (or multiple methods of research) Four evaluation criteria for qualitative research: Credibility, Originality, Resonance & Usefulness Observational research Systematic observation and recording of behavioural patterns Emic perspective Insider's perspective Natural group (focus group) is... A group of people who already know each other Leading questions... Have the answer given in the question Closed questions... Have a fixed range of possible answers Delphi procedure Repeated interviewing of experts until a consensus has been reached Phenomenology Nature of being Ethnography Cultures, customs, habits and mutual differences Grounded theory Analysis and theory development happens after you've collected the data Logistic regression Describes relationship between binary variables (?) Occam's Razor Answer with least assumptions possible Multiple Linear Regression Linear regression with multiple explanatory variables BoxCox Transformation Predicts lander that best transforms the data Backpropagation is used to... Adjust the weights of an ANN based on training data Height of dendrogram represents level of dissimilarities