Thanks Pablo for an interesting comment (so I voted up it up ;-). The dilemma we're facing is that we should absolutely not be chasing RG Score by adding questions and answers left and right (and I unfortunately, I see a fair amount of pretty meaningless answers in particular), but at the same time, this should not impede our discussion activities, and inhibit us from "liking" good answers and questions. Noting how my own RG Score has evolved over the last few weeks, I realize that it has nothing to do with my reputation as a scientist - my numerous publications basically count for nothing - but has everything to do with my activity on the site. So the RG Score is nothing more than rather meaningless activity measure, and it would be good if the creators of ResearchGate would clarify this, instead of claiming it to be a measure of reputation, as this may simply mislead a lot of people.

Dear Professor Bjorn,

Although I agree with you, I think the aim of "this thing" is a bit different. While in real science world we all know who are the big guys, we hardly have the chance to interact with them. Here however, we have the chance to interact , know a bit more in an informal way, and may be even end up collaborating in some "real research". I see what you pointed out, and like you as well, I do not see any sense in increasing your "RG score" by asking, answering and following people as crazy. Although, I am not a big fun of this thing, i think it offers an interesting place for talk about science in a relaxing way. Furthermore, aside from wasting some time, you would never loss you real scientific reputation for interacting a bit here....I hope!.

As I said before, as a forum, I think RG is great. It's just the claim of measuring scientific reputation which doesn't make sense. Maybe the RG Score should be called "curiosity index" instead ;-) (and I, of course, raised my own RG Score by posting this question ;-)

Dear dr Bjornsson,

let me first start saying that I agree with you. More in detail, I disagree with what I think is an increasing trend in building any form of indicatars of "scientific excellence/reputation". I have not authored as many papers as you did. Nevertheless, after 30 years I think I have a convenient and respectable scientific production. Now, the problem is, in my opinion, to agree first on what scientific repution means and, second, to decide if we personally care for the kind of reputation we may gain from this metrics. Besides, once you have accepted the game of the metrics, then what's for? I wonder if one can logically set up an effective tool that can objectively and with the less error possible distinguish among different levels of scientific reputation. To me this is the real non-sense. As a matter of fact, daily (ordinary) research products have to be evaluated time by time with reference to the specific context. I mean, considering their value with respect to the topic, to the local conditions, to the purpose (i.e if they have to be evaluated for a grant application, a job application and so forth). There are no doubts that some, exceptional works, are really above the average, but they are few and, thus, they are excellent by definition and authors gain a good reputation. On the other hand, I think that I DO HAVE a good reputation among my peers, although any metric that I can use to quantify the level of this reputation may provide any answer, yes or no, irrespective of my production (which is a constant in the equation) depending upon the ingredients used to build up the equation. just take a lohh to the H index. When one moves his/her first steps, usually will not immediately appear on top journals. Thus, the number of citations will be presumably low or very low. Then, he/she will improve. As a consequence the H index computed over the whole production can tell us that that particular researcher has no reputation, while the 10-years H index (that takes into account only the more recent production) will tell us that the same person is an excellent scientist. There is obviously something wrong. This is why I really don't care about any reputation indicator. This also why I also agree with the answer from Mauricio Urbina: this is a place were to discuss and exchange ideas without pressure of the kind publish or perish. It is an easy place. That's all.

My best

Dear Francesco,

Thanks for expanding the discussion to the wider question of what is meant by scientific reputation as well as pointing out the ever increasing use of various metrics to weigh scientific output. I think that these are important discussion points. I think it's obvious that journal impact factors, citation frequency of papers and the H-index are increasingly being used in evaluation processes, both for academic positions and research funding, and the question is how far this trend will go? At my university, publication bibliometrics are e.g. now being used to divide some of the basic governmental funding between different departments at the faculty of sciences.

I must admit that I find the H-index rather ingenious in its simplicity of combining the number of published papers and their citation frequency. I have at least 38 papers which have been cited 38 times or more => H=38. For the H-index to increase by 1, at 39 of my papers need to be cited 39 or more times.

This raises the question if citation frequency is necessarily a good measure of scientific excellence? I have a bunch of papers which I consider quite good, but have not been cited that often, and vice versa. Also, I've published papers in top journals which are not heavily cited, while some others, published in lower ranked journals are frequently cited. I draw the conclusion that journal impact factors are overrated (and overused), at least when they're being used to evaluate the scientific excellence of individual scientists.

Cheers, Thrandur

Thrandur, let me attract your attention on a few additional minor points.

Let me also state that I have a bias due to my italian blood (and I do not want to bother here anybody with the present italian academic situation)

First, it seems that the evaluation methods based upon metrics are progressively abandoned by their original proponents, provided that they have not worked as expected.

Second, I strongly suspect (and this is way I have a bias, probably) that in a crowded and yet strong research group I can cite your papers provided that you are citing mine. Not that fair playing ...

Third, a crowded and yet strong research group , yet your direct competior on funds, academic positions and so forth, decides to ignore your papers. Academic killing. Again ...

I remain convinced that a commission has to READ the papers, freely selected by the candidate, and then judge afterward. Obviously a very narrow selection, maybe no more then three.

Then, it is also convenient to evaluate the global production and whatever is thought to be necessary not to hire or support a, let me say, lazy boy.

A good reputation is based upon intellectual honesty and ethical behaviour, that's why I am so sceptical on any form of metrics as an absolute and objective evaluation of reputation. provided that I agree that H index is a good indicator.

Francesco

Hi again Francesco,

It's interesting to hear that you that you're seeing a decrease in use of metrics in evaluations in Italy, and I agree that the most fervent proponents of the use of metrics are losing steam a bit because these are simply so blunt and imprecise instruments. However, I sit on the board of the appointment committee for professors and lecturers so I both read the applications by the applicants, and also read the external evaluations we request from both Swedish and foreign expert evaluators. It's interesting to see that some evaluators place a lot of emphasis on the "metrics" while others don't. However, my feeling is that it's still being more and more used, not less. Of course, all the evaluators are also requested to read the selected papers that the applicants attach to their applications and also make their evaluation based other information in the CV.

Your comment on biased citation practices (especially now that citation frequency is being used as a measure, e.g. in the H-index) is a valid one, and I think it's both a universal problem and has always been there - research groups that predominantly cite their own papers and ignore papers by other groups in the field. I agree with you that this is a question of fundamental scientific integrity.

Well, tomorrow I'll be in Rome for a weekend break, and am very much looking forward to it!

Cheers, Thrandur

Well, enjoy your stay!!! Just to say, I meant not in Italy, but elsewhere... Going hom now ...

Thanks Pablo for an interesting comment (so I voted up it up ;-). The dilemma we're facing is that we should absolutely not be chasing RG Score by adding questions and answers left and right (and I unfortunately, I see a fair amount of pretty meaningless answers in particular), but at the same time, this should not impede our discussion activities, and inhibit us from "liking" good answers and questions. Noting how my own RG Score has evolved over the last few weeks, I realize that it has nothing to do with my reputation as a scientist - my numerous publications basically count for nothing - but has everything to do with my activity on the site. So the RG Score is nothing more than rather meaningless activity measure, and it would be good if the creators of ResearchGate would clarify this, instead of claiming it to be a measure of reputation, as this may simply mislead a lot of people.

Now they're flagging for a change in the RG score. On your dashboard it now says: "The RG Score is changing! We're about to incorporate more data into the RG Score. Check back soon for the update". "It's getting curious and curiouser" as it says in Alice in Wonderland ;-)

Thanks Shweta for making a number of good points. Most of these underline the basic value of RG as a forum for young and old scientists alike, with or without a RG Score. The recent change made in the calculation the RG Score certainly make it less nonsensical than it looked to me when I originally posed the question.

So I do see an advantage of the RG Score for your second point; if someone wants to increase their RG Score by attracting discussion and "upvotes" to your questions, they need to be timely and well formulated.

Dear Amy, Yes, this is an "old" discussion, but it's still highly relevant, so it was interesting to see your reflection on the subject. I raised the question when RG introduced the RG Score (or at least when I came aware of it) and my immediate response was "what's this nonsense"!? The reason for my rather flippant question was the fact that RG was telling us the this was a new measure of our "scientific reputation" when it's really NOTHING OF THE SORT! (sorry for raising my voice ;-). When it comes down to it, the RG Score is a simple (and really very banal) activity meter.

For the makers of a social platform, this is a psychologically very clever way of keeping the users interested, similar to the "likes" on FaceBook etc. In order to be "liked", you need to be active.

So, for awhile, I argued that the RG Score should be removed, but then there were a few young colleagues that pointed out that for them, with no or few publications, this was a way to show that at least you are active in the discussion on RG. If you post questions or answers only to increase your activity, and these as stupid or irrelevant, they should be down-voted, so that you should not be able to gain RG Score by posing dumb questions.

I see my own RG Score go up and down a bit from week to week, and it has all to do with activity on RG, not so much about how many publications I have, certainly not how often they are cited (both highly relevant to the "reputation" you have as a scientist) and neither does the number of your "followers" enter into the equation, it seems.

So I just take the RG Score for the nonsense it is. However, what you point out is quite serious, and that is if academic evaluation committees, evaluating job or grant applications, are using the RG Score as a scientometric variable in their evaluation. Then it becomes a discussion similar to that about Impact Factors and H-index. What does these values really mean, when is it relevant to look at them, and when are they simply "nonsense".

In my mind (I haven't done this yet, though), the only time I might think the the RG Score is relevant, is at the undergraduate level, to find out if a student seeking to do an undergraduate research project in my lab is showing enough active interest in science in general to be on RG and posing answers and questions (and then I could see the quality of those answers and questions).

Cheers, Thrandur

I'm new to this RG lark, and it is all rather mysterious. My RG is 99% based on my publications; whcih would make sense as i've only recently joined. What's the Impact Points? Is that more informative or also heavily manipulated by joining in the answer/question game?

I joined a while ago and my RG score is still 100% made of publications, or so it appears, probably because I did not participate in discussions. Does anybody know if there is a description anywhere of how the algorithms work - how does citation of papers are taken into account, impact factor of a journal, "skills endorsements" (!) or anything else. Everything seems to be locked in a black box. What the hell is Impact Points and how are those different from RG score?

A clear rules of the game should be spelled out and to be known to the individual players.

Hi guys. I have had an RG account for a while and generally ignored the RG score as it means nothing to me. As Bjorn pointed out, h-index IMHO is a more established parameter to indicate the quality of a researcher. However, I got an email yesterday telling me that my RG Score has been decreased by 0.01. Although this is negligible, it just makes me uncomfortable that my score was decreased. So I have been penalized for something that I did or didn't do... It just drives me nuts... If this is their way of getting me to pay more attention to the RG Score, it's working...

90% of my score is publications...I think they have fixed this problem. We can agree that asking questions and helping other with their questions is important and is not measured in raw publication numbers at all. This is SOCIAL media after all.

Totally agree, questions and answers should be part of the scientific reputation of a researcher.

The points of discussion here is very interesting as I have always been asking this question to myself. What is the RG score and how is it calculated.

Dear Professor Björn Thrandur Björnsson, 

Beyond the score, RG has sense of bringing researchers on a platform to discuss.

RG score is totally confusing tool, no idea how it works....? i have gone through many articles of very high impact factors and citations, but authors have earned very low RG score compared to those with 1-2 articles with low or no impact at all but gained high RG score.

I sincerely hope that the up or down votes on answers play a part-- I have seen many people who admit they do not know the answer to the question posed, yet go on to answer some other question.  The only reason to do this would be an attempt to inflate your score.  I wish there were a button for exempting a posted answer from the RG score, as I would use it on the one you are reading (mine).

I think it should be stated somewhere how RG score is calculated

You can visit https://www.researchgate.net/publicprofile.RGScoreFAQ.html for a brief explanation of RG score 

There should be known measuring sticks. Grouping also influences in positive way but its painful when when it adapts a negative approach for a good research work or for a exact and to the point answers. 

RG-Score has few evaluating components whose weights are not clear. How to check the negativity is also not known which may turn to a dangerous point for RG. 

We are playing a game whose game rules are not spelled out, and that is RG score.

I think, we all should know how RG Score is calculated.

Joined RG not very long and I am perplexed as confused by this so-called RG-Score. Would anyone have an FB-Score for Facebook? I mean every cutex post on Facebook should earn that community a FB-Score? I thought this was a community forum to interact with other researchers and not necessarily look for scores.

Dear Dr.  Björn ,

I don't know and this creates great expectations!

But in anyway RG is providing that I know researchers from all the World and can get in contact with everyone !!

Best Regards,

Vanessa

@Kato Mivule. Althought there's no "FB-score" a lot of Facebook users are extremely focused on number of followers, number of likes and such, the same with Tweeter; number of followers, number of retweets... So even in there's no "FB Score" I think most all social networks apply various numerical measures - and for some reason this appeals to people.

@ Vanessa Meloni Massara. You focus on the positive aspects of RG and I'm in full agreement with you!

Yes it is true that RG provides enormous benefits but our discussion is about RG score.

RG score is the number of questions and many other activities but to evaluate scientific reputation i don't think so.

As I had witnessed as well, the RG score is absolutely a dump. Hungreds of friends had downloaded and viewed my recent publications. Yet there is no increase in the score whatsoever..... So I see its nonsense.....

I have 0% of RG Score and I has asked and answered questions, I has published my research and follow people and I still with 0%. What can I do?

Hi Héctor, Lucky you!  Now you don't have to worry about this silly RG Score ;-).  Apart from that, if you have joined RG quite recently, there may be a lag-time (if I remember correctly) until your RG Score shows up for the first time. 

Thank you, probably it is because I have less than one week here.

Hi again Héctor. That definitely explains it. Your RGS will appear in a week or two, I'm sure. And if you really want to play this social media game, don't forget to up-vote answers (and questions) you find useful, and equally, to down-vote those you find bad, as this will affect the RGS of your peers on RG. 

you should not be worried about the silly one.

The RG score means the researcher reputation and participation in the RG community. The essential is the IP impact point reflects the impact of scientific work of the researchers.

Thanks a lot of this question and valuable answers

I completely agree with Björn.  RG score is simply a way to keep users interested and active.  Academia.edu does something similar in the form of 'top 2% by 30-day views".  The way Academia.edu does this seems more legitimate however.  First of all, the 'top 2%' or 'top 3%' is based on the number of people who have downloaded one's papers.  One can actually see which papers are the most popular and earning the top marks.  But with RG score, I am baffled.  I added a lot of publications to my ResearchGate profile recently and my RG score didn't budge.  Then, it went up by 1 point, for no apparent reason.  Soon after that, I requested a paper from someone with a lower RG score than mine.  That person sent it to me then started following my profile.  My RG score immediately went down.  That makes no sense.  So, I wouldn't take RG score too seriously.  Simply use ResearchGate to interact with colleagues and find interesting work that other people are doing.  And, I'm not posting this to raise my RG score, by the way :-)

I have seen multiple questions like this on RG. I can only contribute that I believe you can only raise the RG based on papers that do have an impact factor "evaluation". The impact points is just sum of all paper's impact points (if journal title is entered correctly). 

There are some questions which are very closely connected to one raised here:

1. Can any single number ever be representative for all your accomplishments? 

2. Do we really need another measure of reputation compared to impact points? (Not evaluating the benefit of impact factor here)

But I think the most important one is: 

3. Why does RG not publish the algorithm behind the RG score? 

a) Hypothesis 1: They still are changing the way it is calculated, thus it is not ready for being revealed, yet

b) Hypothesis 2: They do not want to publish it, because they have not put any scientific meaning into the algorithm (makes the RG kind of random)

c) Hypothesis 3: They just want to keep people interested in threads like these. 

To end this, I think it is time that some (influential) group of users, write an open letter to RG to ask them to provide the algorithm for the RG score. Only by then, one could argue if it is "non-sense" or an actually well-pondered, mature concept. 

So I highly recommend, that you, who seems to be highly interested in this topic initiate something like this to finally stop all the wondering about RG score.

Good response Matthias, I am also curious about the algorithm behind the score! I know the score is largely based on the number of publications and their impact factors, but it would be great to see exactly how it is calculated and how the different contributions are weighted.

Hi Matthias,

I agree that the "impact points" score on RG is probably the sum of the IF of indexed papers (even if it didn't exactly add up when I tried adding up your IFs ;-)

However, the RG Score is only affected to less extent by additions of papers, as it appears mostly to be based on RG activity; Q&A and up/down voting.

I can subscribe to all three of your hypotheses :-)

Here below the RG man (I think he knows the RG score calculation)

https://www.researchgate.net/profile/Joachim_Pimiskern/reputation

And If you visit his profile the RG score will get up more and more ...

Hi Alessandro,

I don't how if he knows how the RG score is calculated, and I don't think it rises when I visit his page. However, it was entertaining to do so. He appears to be a human Google interface - anybody who poses a simple question on RG, he Googles the answer and gives the web link. LOL!

Dear Bjorn, yes .... you discovered the formula !!! A software interface can easily overcome everyone's RG score grasping the best from google and wikipedia.

RG company should ask to Google team to score the questions and answers as inversely related to web info redundancy. 

Hi Peter.

You make a good point about the strategy of posing questions vs answering them. This can be affected by us by actually using the up- and down-voting options. Down-vote stupid questions, up-vote good answers.

Regarding the "RG man", he's following even a greater number of people than are following him, so he may be working on the principle that if you start following somebody, that person will automatically click on the "follow back" button. If that's his strategy, it seems to work about 70% of the time. 

I also have absolutely no idea what non sense is this RG score. I have been using Research Gate for a long time and has got publications listed here still my RG score is inactive. This has made me understand that one should not consider this score anything to do with a persons scientific reputation. 

In fact, the effect of this seems to be quite high.

Dear Hosny,

I'll follow your tip and will ask few questions. As far as the participation is concerned, I have been replying to several questions posted by other researchers. Anyways thanks for the advice.

Dear Prof. Björn Thrandur Björnsson,

I agree that a clear rule about question answer RG-score is not real reflection of our strength.

Just wait a week and it should be there!

You will, I am pretty sure. :) It may feel like it takes strangly long in the beginning, though.

Hello,

I was initially very skeptical about the RG score; Seemed like another competing "construct". Yet -- after closely observing its behavior, I am increasingly more comfortable with the RG score actually being a valid measure of scientific reputation. Below are some examples to support my statement:

1) RG score does take into consideration one's "Impact Points"; Yet, all of us know that impact factor (which the Impact Points are based on) is flawed and does not accurately reflect one's true contributions to science. An example here might be a world-changing book that has absolutely no impact factor assigned to it, or perhaps a scientific conversation on one of the RG fora that ends up in a significant discovery...

2) Participation in discussion boards on RG allows exchange of information to occur at a level previously impossible; For example, a Nobel Prize winner may be much easier to reach on RG than in real life... etc.

3) ResearchGate rewards uploading of original data; Thus, datasets that previously would likely "fade from the pages of history" now can be shared, re-analyzed, consolidated, and voila... new hypotheses generated!

4) Finally, the RG score provides an objective measure of scientific participation; Members get to up/down vote answers they think are more valuable or "useless" while the discussion evolves and continues... although this may seem a bit childish, it's actually quite powerful --> The best answers tend to eventually "float up"...

There are more points I could make, but I think the ones listed above support my case the most. It is always easy to provide constructive feedback, but I was always guided by the principle that before one provides criticism, one ought to have a working solution for the problem at hand.

My five cents!

Regards,

Stan.

" Deleted" · University of Science Malaysia

Hm...

I don't think that RG Score has a relevance in the academic world outside RG. No need to give too much consideration to RG Score. Please take it as an indicator of your activities in RG. What the advantages of RG is it really helps to get remarks from experts in relevant fields and to popularise our research articles.  No further consideration is necessary related to RG and RG Score.

I guess that we will have to wait: new metris always take a while to spread. Changes are good, though, that RG score will be superseded by just another metric before it gains broad relevance...

Dear Prof. Björn Thrandur Björnsson,

I think the RG score evaluation process has modified.  I was looking my RG score statistics and found the following note. 

"Please note: We are currently working on enriching publications with more data. You may notice a few changes to your RG Score as a result."

I agree with your latter posts Björn that the RG score is an acceptable metric provided people understand what it represents. I can see the value in a score that measures engagement and interaction with colleagues but it shouldn't then be extrapolated to mean more than what it is. Perhaps this is why the definitions have shifted away from directly referring to 'reputation' and now state that the RG score is to "measure and leverage your standing within the scientific community".

There is no secret to the factors that contribute to the calculation of the RG score. If people know how a score is derived then I hope they are able to interpret it appropriately.

As Afaq points out, RG is clearly developing it's calculation of the RG score and it will continue to change. This is the current information  RG gives on this process:

So, what’s next?

As it’s been designed to adapt to the needs of the research community, the RG Score is constantly evolving. Here are some of the updates you can expect in the near future:

Evaluation

We’re currently working on ways to give you more options when it comes to evaluating your peers’ contributions.

Disciplines

We’re adapting the RG Score to include a lot more discipline-specific data."

Richard. I was happy to read your response telling me that RG had "shifted away from directly referring to reputation...", but unfortunately, I find this not to be the actual case. When I click on my "RG Score" link, it says "A new way to measure scientific reputation" and when I click on "Learn more" about the RG Score, you're met with the statement:

The RG Score is a metric that measures scientific reputation based on how all of your research is received by your peers.

The RG Score is, of course, however the people of RG try to define it, just a bit of trivial statistic, and I cannot be but slightly amused when I see that my 200 papers written over a 30+ year career make up 36% of my "scientific reputation" according to RG, while the answers I've provided on RG (which all in all have probably taken me a good few hours to write) account for 42%. :-)

The last thing that you mention is probably the problem with most metrics. Like, there are some people who hardly write papers but have had a seminal book published two or three decades ago, and are having a high reputation by any degree.

I totally agree that what is happening inside RG is probably overrated by RG. This is not surprising after all, though. ;)

It is the first time for me to hear how RG scores come from...

For me, the RG score is still inactive. I don't know when it will be active. What a complicated process :)

Me too, Arnat

Odd...particularly by adding indexed publications it should be there, soon.

Maybe,wait  for it

me too my score is currently inactive i dont know why or whenbe  active agian and how

Was it active before?

yes it was cative before

Now that's strange. Did you delete papers?

What bad effects does it have after deleting a paper?

I do not know for sure, but since adding papers usually increases the score slightly, it should decrease it slightly.

Thanks, I think so,,, I accidentally deleted one paper...

https://www.researchgate.net/publication/275521543_A_relational_altmetric_Network_centrality_on_ResearchGate_as_an_indicator_of_scientific_impact_Journal_of_the_Association_for_Information_Science_and_Technology

Article A relational altmetric? Network centrality on ResearchGate a...

 Good article, Vuan

Hehe, it was just a question of time.

I'd like to rank with a positive score scientists that shut down computers for one or two days per week, putting their glasses for reading papers and books printed with a small font size. 

 Thanks Alessandro Vuan for drawing our attention to this paper analyzing RG in terms of scientific impact.

The paper is interesting, but as the authors themselves point out, the data they analyze are very limited, based on 55 persons from a single academic department. In my opinion, RG is little used for within-department exchange as most departments will have physical proximity between faculty members and a lot of well-established instruments for scientific interaction. 

Another concern is that the data used is from 2012-13 so they are already somewhat outdated in terms of how RG calculates the RG Score, although the underlying data are still valid.

Lastly, this type of analysis needs also to take into account what scientists and why, are NOT active on RG. In my mind, these often include the most established and influential scientists as a) they're in their 60-70's and perhaps not so active on ANY social media, b) they have many other channels to impact on the scientific community.

Besides, much exchange is happening via non-public channels, e.g. as part of project work using groupware...

The Introducing RG Score page suggests two ways: "To get started, ask or answer questions in Q&A, or publish your research to your profile." I did the second way but my RG score is still inactive. This guide need to be changed. 

Btw, how to have the Impact Points and have it displayed on RG? Some brief information should be introduced somewhere near the introduction of RG score.