Close Home Forum Sign up / Log in

Understanding H Index and High Impact Factor

T

Hi,

I asked this particular (new) librarian and according to them, there are two ways academics measure "quality";
1. High Impact Factor
2. H-Index

As I understand it,
1. High Impact Factor:
Published by Thomson Reuters which produced the ISI Web of Knowledge DB. The DB has the Journal Citation Report (JCR); JCR will show the high impact factor result(s). According to the librarian, it calculates the overall journal impact, not the specific article(s). It's only used for journals (?). I'm not sure if it's peer reviewed.

2. H-Index:
Produced by Scopus and is peer reviewed. Use H-Index to calculate the article's quality, as well as the writer of the article (? - not sure about this one)

Both will not include areas such as religion, (specific) languages and history.

Ok, I may be wrong; could someone help me out if I did? I would like to understand them more

H

The 'impact factor' is a measure which is applied to journals. It reflects the average number of times that papers in that journal have been cited in a specified period (last two years?). For example, Journal X has an impact factor of 2.54, meaning that on average each of the papers published in it have been cited an average of 2.54 times in the last two years. Which of course means that some may have been cited a lot and some not at all. It tells you nothing about the quality or citation rate of an individual papers within that journal.

The 'H index' is a measure which is applied to individual researchers. It is a slightly odd concept - it tells you the number, n, of that person's publications which have been cited at least n times. For example, I have my name on 4 publications of which one has never been cited, and three have been cited at least 3 times, so my H index is 3.

Both of them have their weaknesses, are subject to gaming and misinterpretation. For example, you can get rubbish papers in high impact journals, or vice versa. Equally, if you had someone who had published 10 papers, 8 of which had never been cited, but the other two papers had been cited 500+ times, they would still only have a H index of 2. So, in my opinion, they're kind of over-rated, but we're stuck with them.

T

Hello, thank you for the info but I have a question: is the H index the only way to measure the quality of a particular researcher? and, where can I find it?. Besides, is there an index to measure the quality of an individual publication?
Thank you again

T

Quote From HazyJane:


The 'H index' is a measure which is applied to individual researchers. It is a slightly odd concept - it tells you the number, n, of that person's publications which have been cited at least n times. For example, I have my name on 4 publications of which one has never been cited, and three have been cited at least 3 times, so my H index is 3.

Both of them have their weaknesses, are subject to gaming and misinterpretation. For example, you can get rubbish papers in high impact journals, or vice versa. Equally, if you had someone who had published 10 papers, 8 of which had never been cited, but the other two papers had been cited 500+ times, they would still only have a H index of 2. So, in my opinion, they're kind of over-rated, but we're stuck with them.


Thanks HazyJane.

Ok, after reading some more from http://goo.gl/8T0qG:

"‘The index h, defined as the number of papers with citation number greater
than or equal to h, is a useful index to characterise the scientific output of a
researcher.’ JE Hirsch"

also from http://goo.gl/FyQgH

"a h-index of 20 means the researcher has 20 papers each of which has been cited 20+ times."

Assuming that I have 7 papers and the number following it is how many times it has been cited:

paper 1:10
paper 2:12
paper 3:15
paper 4:15
paper 5:14
paper 6:1
paper 7:0

My H here will be 5.

But if:

paper 1:2
paper 2:2
paper 3:2
paper 4:1
paper 5:2
paper 6:1
paper 7:200

Is my H 1?

H

@tt_dan In your first example, you're right, the H index is 5, but in the second example the h index is 2, because there are more than 2 papers that have been cited at least twice. If paper 1 and paper 2 got an extra citation each the H index would go up to 3.

@Tonypane - the way to assess the quality of a specific paper is.... read it. That's pretty much all you can do. In some fields there are guidelines as to what a 'good' paper on a particular study design should contain e.g. how to write up a clinical trial. But at the end of the day the only way to tell if a paper is of good quality is to read it and apply the knowledge and critical reasoning you have developed through your own studies.

If you're looking for H-indices of particular researchers, these can be found on people's Google Scholar profile pages if they have one. It might also be available via ResearcherID.com.

In general... don't forget that a paper can be cited for the wrong reasons (lots of people critiquing it for example). And there may be excellent papers that get a cited a couple of times but have a big real world impact on say, policy. So these are very approximate measures, and I think we shouldn't invest too much energy into them.

T

Quote From HazyJane:
@tt_dan In your first example, you're right, the H index is 5, but in the second example the h index is 2, because there are more than 2 papers that have been cited at least twice. If paper 1 and paper 2 got an extra citation each the H index would go up to 3.



Hazy, I think I'm confused : (
I counted the first one as 5 because there are (exactly) 5 papers with more than 5 citations, so this fits the definition of: "h, defined as the number of papers with citation number greater
than or equal to h"

For the second one, at first I counted 1 because the highest citation is 200; but I don't think we're counting the highest number of citations.
Are we counting the number of papers with the largest recurring citations?

I'll give this example:

e.g.
paper 1:11
paper 2:11
paper 3:11
paper 4:11
paper 5:12
paper 6:12
paper 7:12
paper 8:12

So, for the above, is the H: 4?

paper 1:3
paper 2:3
paper 3:2
paper 4:1
paper 5:2
paper 6:1
paper 7:200

Going back to the example, why is H 3 here, since there are only 2 papers that have 3 citations

H

Google Scholar defines the h index as "the largest number h such that h publications have at least h citations." This means that the h index will always be smaller than the total number of papers published by that person.

In your first example the h index is 8 because there are 8 papers which have AT LEAST 8 citations.

In your second example there are 3 papers that have AT LEAST 3 citations.

My H index is 3 cos I have a total of 4 papers which have been cited 0, 3, 15 and 30+ times respectively. So there are 3 papers that have been cited at least 3 times.

If you had to have an exact number of papers with exactly the same number of citations the number would fluctuate up and down too much as citations increased, whereas it should actually increase over time for an individual, assuming their citations were increasing.

Hope that clarifies things. Like I said, I really don't think it's worth getting too bothered by. I for one wouldn't know how to compare those two candidates on the basis of the h index. I'd want to read their papers and know what their personal contribution was.

23968