Age, Media and the Assessment of the European Union
INSTRUCTIONS:
• IMPORTANT:
A. The interpretation and description of the results are essential!
B. Prior to the analysis the variables MUST be “cleaned” from missing
values and/or non-responses!
C. This is not a group work and you are expected to work on this assign-
ment individually.
• _ Write your results in a document (.doc) renamed as “Surname”.
• _ Make sure you include all Stata commands.
• _ For this homework we will use the usual AUTNES data ZA5859.dta.
• _ DEADLINE: February 26, 2015 at 11am.
? Download and load the data.
For this series of questions, we will use the usual AUTNES data “ZA5859.dta”.

1. Recoding of variables.

Create a dummy variable called age_new that takes a value of 0 for the respondents
aged 20 or less and a value of 1 for those aged 70 or more. All the other observations
should be set to “missing values”. (1 point)

2. Table Analysis.

Examine whether or not there is a connection between groups of respondents’ age and
their media behavior by answering all the questions (Q2a(i, ii) and Q2b (i, ii)).

(a) Is there a connection between age groups (age_new) and the consumption of
newspapers in general (w2_q36)?
i. Formulate two hypotheses (a null hypothesis and an alternative hypothesis) for the relationship between age groups (age_new) and the consumption of newspapers in general (w2_q36). (1 point)

ii. Choose an appropriate statistical test (a Chi-square test OR a Spearman’s test OR Person’s correlation test OR etc.) and examine whether or not
there is a statically significant relationship between age groups (age_new)
and the consumption of newspapers in general (w2_q36). (1 point) [ADDI-
TIONAL INSTRUCTIONS: here you need to discuss whether or not based
on the results of the chosen statistical test you decide to accept or reject the
hypotheses stated in Q2ai].

iii. Explain the results of your test to someone that knows nothing about statistics. (1 point) [ADDITIONAL INSTRUCTIONS: this means that you can not use words like p-value, null hypothesis, etc.].

(b) Is there a connection between age groups (new_age) and the frequency of con-
sumption of Krone Zeitung (w2_q37x6) specifically?

i. Formulate two hypotheses (a null hypothesis and an alternative hypothesis)
for the relationship between age groups (age_new) and the consumption of
the newspaper Krone Zeitung (w2_q37x6). (1 point)

ii. Choose an appropriate statistical test (a Chi-square test OR a Spearman’s
test OR Person’s correlation test OR etc.) and examine whether or not
there is a statically significant relationship between age groups (age_new)
and the consumption of the newspaper Standard (w2_q37x6). (1 point)
[ADDITIONAL INSTRUCTIONS: here you need to discuss whether or not
based on the results of the chosen statistical test you decide to accept or
reject the hypotheses stated in Q2bi].

iii. Explain the results of your test to someone that knows nothing about statis-
tics. (1 point) [ADDITIONAL INSTRUCTIONS: this means that you can-
not use words like p-value, null hypothesis, etc.].

3. Comparison of mean values.

The question w1_q26x12 asks respondents whether they 1. “completely agree”, 2. “some-
what agree”, 3. “neither agree nor disagree”, 4. “somewhat disagree” or 5. “completely
disagree” with the statement: “The Austrian government should stop immigration to
Austria”.

(a) Using the ttesti command check whether the mean age of the respondents
giving a score of 1 to the question on immigration (w1_q26x12) is significantly
differ from the mean age of the respondents giving a score of 5 to the same
question. Remember that the variable that stores the information w1_q26x12
must be cleaned from missing values prior to analysis. (2 points)

(b) Discuss the results of t-test. [ADDITIONAL INSTRUCTIONS: here you need
to discuss whether or not based on the results of the t-test you conclude that
the two groups differ significantly in their assessment of immigration.] (1 point)

4. Recoding of variables and table analysis.

(a) Now build a dummy variable called new_kronen indicating whether a respondent
reads the newspaper Kronen Zeitung often (4 days a week or more often) or rarely
(less than 4 days a week or never). The variable needed is w2_q37x1 and must
be cleaned from missing values prior to analysis. (1 point)

(b) Examine, using an appropriate method, whether these two groups of respon-
dents (new_kronen) differ significantly in their assessment of the European Union
(w1_q22). (3 points)

5. Interpretation and evaluation of the results.

Now interpret the results of the above exercises answering the following questions.
Interpret the results in layman’s language so that anybody, even one from a non-
statistical background, understands them [ADDITIONAL INSTRUCTIONS: you must
not use terms like statistical test, chi-square, p-value etc. In addition, you must use
at least 80 words for each of the questions 5a-5c].

(a) Is the assessment of the EU related to age? Do younger people hold more positive
attitudes towards the EU? (2 points)

(b) Is media consumption related to age? Do younger people read newspapers more
often? (2 points)

(c) Is media consumption (specifically the use of the Kronen Zeitung) related to the
assessment of the EU? (2 points)