Professional Documents
Culture Documents
Project IBM
Project IBM
Deadline: 15.01.2020
P1. Make a statistical research, based on a sample from a population, including variables
related to each other. Should be included at least two/three variables from each type,
respectively: three quantitative variables at least two being continuous, qualitative -
nominal data, qualitative - ordinal data. The collection of data will be made with the aid
of a small questionnaire, or from other sources (Eurostat, World Bank, …);.
Your sample has to include at least 40 units.
Make a short research’s presentation: statistical population under study,
elementary unit, the research’s objective, the sample size, type of sampling, describe the
variables.
P3. Choose the main quantitative variable X from your statistical research (connected to
the investigation’s objective):
a) compute and interpret the confidence interval for mean of X, maximum error of
the estimate E, standard error for the mean S.E.. Analyze/Descriptives
Statistics/Explore. From Statistics, Descriptives.
P5. a) Select two qualitative variables: i) Obtain the crosstab, containing the observed
and also the expected frequencies (counts); ii) Test if the 2 variables are independent of
each other, using the chi-square statistic; iii) Study the degree of association between
variables (Analyze/Descriptive Statistics/Crosstabs, from Statistics, Cells).
b) Select 2 variables: a dependent variable X which is quantitative, and an independent
variable C (factor) qualitative. Perform the Analysis of Variance (ANOVA): i) What say
the statistics from Descriptive and the Means plot (from Options)? ii) Compute F calc
and decide if the factor C has a significant effect on dependent variable X.
Analyze/Compare Means/One-way ANOVA; Options.
P6. Data consist of a sample of countries (at least 20) observed relating to 3 variables: a
dependent variable Y and 2 independent (explanatory) variables suggested by the
economic theory; data should be downloaded from World Bank or Eurostat. a) Estimate
the multiple linear regression; b) Decide if each of explanatory variables is useful in
explaining the variation of Y (see t and Sig. for Coefficients B) and maintain only
variables with significant effect on Y; interpret the regression coefficients and obtain the
confidence intervals for these; c) Compute and analyze the multiple correlation
coefficient (R). Apply F-test (from ANOVA table) and decide if the model explains a
significant part of the variation in Y; d) Obtain predictions and confidence intervals for
predictions; e) Introduce a dummy variable in regression, interpret the coefficient of this
variable and decide if the dummy variable contribute to explain variation in Y; f)
Estimate also a nonlinear regressions, suggested by scatter plot, considering only one
independent variable; (Analyse/Regession/CurveEstimation).
Observations
A. The project should be written as a word document (with outputs saved from SPSS);
the data bases will be delivered on a CD. This is an individual work, each student has
own database. If the project is delivered after deadline, the grade is diminished with 1
point (ex. from 10 to 9).