We have come up with 19 questions, and respective answers, enabling building different tests by sampling 10 questions each time you request for a new test.

We are not recording any results. If you want to keep a copy, please print the test before taking a new test or leaving the page.

Taking the test has no influence on your admission to the Tutorial. It has been designed to complement the information provided in the description of the tutorial.

We recommend you reading the document “Basic concepts” before taking the test. The document is available in the Supporting material section, .

We hope you enjoy and learn from taking the test…



1.

Associating addresses to wrong locations for whatever reason may add difficulties during the unification, agreed. But how do you think it would affect a ranking of countries, for example? Try selecting all the correct answers.

a)
None visible effect! If addresses have been located to other places inside the same country.
b)
None visible effect! Even If addresses have been located to other countries.
c)
A complete disaster! Locating addresses to a different country would invalidate the ranking.
d)
It may affect the number of publications attributed to the countries in the ranking, but it will hardly change positions in the ranking unless the number of miss-located addresses is huge.
e)
It may affect the number of publications attributed to the countries in the ranking, but it will be really hard to tell if there is any flaw because bibliometric studies are not easy to reproduce.


2.

When the address field is empty there is no possible way to attribute the publication(s) to any organization.

a)
Of course! There is no way to attribute that publication according to the method based on addresses. In that case I would report the fraction of publication not attributed for this reason.
b)
Nope! There is still the possibility to attribute that publication according to authors.


3.

Suppose you have a publication with 4 addresses, each of which link to a different institution. What would be the number of publication per institution according to the fractional counting method?

a)
One publication per center.
b)
A quarter of publication per each of the centers.


4.

Suppose you have a publication with 4 addresses, each of which link to a different organization. What would be the number of publication per center according to the full credit method?

a)
Four publications for each of the organizations
b)
One publication per center.
c)
A quarter of publication (1/4) per center.


5.

Suppose there is a huge old and reputed university in your R&D. The university has campuses all over the country. During the last 15 years campuses in remote regions have developed to the extent that today they are autonomous organizations, each with its own name. As bibliometrician you are requested to describe the evolution and current status of the system based on bibliographic data of the last 20 years. How would you treat this university (concerning attribution of publications of course)?

a)
I would try to capture all the changes of any nature occurred during the period.
b)
I would try to achieve the objectives of the study: describe the evolution and current status of the system.
c)
The situation recreated in this example puts me in a difficult position. I would discuss the objectives of the study with my clients in order to try to find the best and feasible solution to this situation.


6.

Do you think that cities or other administrative units may help in unifying addresses?

a)
Non sense! There are different cities around the planet with the same name.
b)
Most probably, but only if the organization under analysis has a disctinctive feature among the rest organizations in the region.


7.

During the unification some addresses may be mistakenly unified. Do you think that assessing the number of addresses mistakenly linked would be a good indicator of the quality of the unification process?

a)
Yes! But it would apply only to the subset of addresses enabling identifying an institution.
b)
Nope! The proposed method to assess the quality would not take into account publications without addresses, neither those addresses not enabling identifying institutions. It would provide a partial view of the process.
c)
Nope! The proposed method to assess the quality may be sensitive to biases since it is limited to the subset of addresses enabling identifying institutions.


8.

Natural language can be so imprecise when it comes to describing the structure of an organization. Do you think drawing a organization chart may help you having a clear picture of the structure of an organization?

a)
Yes! Especially in the case of large organizations.
b)
Nope! Everyone knows what a university or a hospital is. There is no need to do such thing.


9.

What would you do if an especially informative piece of information of an address, like the name of the organization, is missing?

a)
Try to identify the "mother" organization by other indications like: campus name, schools, location, postal code
b)
Nothing! I will skip the case and continue unifying the rest of the addresses. I have a lot of work to do!
c)
Nothing! I will skip that case anytime I find it.


10.

You have been contracted to carry a study of a research institute composed of different organizations.

a)
You would first arrange a meeting with managers to have a clear definition of the institute. After that you would start attributing publications to the institute.
b)
You would prepare some short reports to show your client how good you are.