Who subscribed to Habrahabr?
Hi, Habr!
A year has passed since my previous article, and my grafomanskaya nature got the better of laziness and made me publish another small study.
The title of the article is, of course, slightly hypocritical (but it sounds beautiful). It would be more correct to call it this way: “Who is subscribed to Habrahabr Vkontakte”.
')
As you probably already understood, it was interesting for me to see who, from where and how Habr reads, likes Habr, repost Habr (however, the last two points in this article did not fit). But at the same time understand what cities, countries, universities, etc. the most "zahabrennye" in VK.
Who cares to see interesting charts, graphs and figures about this, please under the cat.
Warning
First, a couple of warning lines. I am not a professional sociologist, I am not a professional SMM researcher or web analyst. Some things I could overlook, some concepts can be called somehow unprofessional. For this, I ask you not to beat (at least in the face), but to point out in the comments on inaccuracies and comments. As far as possible, I will supplement and correct. Thank.
Also, I draw your attention to the fact that the sample under study is a public audience from the social network “Vkontakte”. This means that user data in it may periodically change, they may be incorrect or inaccurate. Therefore, when I say “Habr's readers are 146% of 91-year-old men from the Isle of Man”, this is not the ultimate truth. Just this is the information specified by users in the profiles.
One last thing: in this article I do not test any hypotheses. This article is rather a review, so there will be just graphs and diagrams, no verification of statistical significance, no regression models, no degrees of freedom and chi-squares.
UPD. 11/18/2014: so it turns out that on many graphs and maps appear the Crimea and its cities. By default, in Google GeoChart it is assigned to the territory of Ukraine, in VK - to Russia. Please do not look for any of my intent in attributing Crimea to this or that state, on all graphs and maps I simply left everything as default and unchanged. The active political position is good, but not in the form of sracha in the comments on Habré.
Okay, pretty chatting, let's get down to business.
Data downloaded using VK Api. They were pumped into a self-written database, where I processed and analyzed data. Initially, I wrote this database for the analysis of other groups and public Vkontakte, but when the work with them was finished, I could not resist and decided to analyze Habrahabr too.
However, this article is not devoted to the description of the technical aspects of the process of obtaining data from Vkontakte. There are enough examples of working with VK Api on Habré, and it’s not worth describing database design and development. If you have any questions - write.
Also, if there are any proposals to analyze some other parameters of users, about which I did not describe in the article or to subject any other public and groups to the same research, write in the comments.
The sample for the study is solid, that is, the entire audience of the public. The data were received in October 2014, so they may differ (most likely, slightly) from the current ones.
To simplify the narration, I will use the words “Habrahabr readers” or “Habrahabr subscribers” instead of the phrase “Habrahabr readers” (yes, in the context of this article, I’ll just say “subscribers”). Just remember that we are talking about the "contact" readers of Habr.
First, take a look at the floor of subscribers:
Chart 1. Distribution of subscribers by sex
The picture is quite logical. Habr is still a site for techies. Historically, there are more males among techies, and therefore there are more males among subscribers (despite the fact that in general on Vkontakte the ratio of men / women is about 53/47).
Let's look at the number of readers by age (I took a range from 16 to 65).
But you will fairly note: so after all, active users of “Vkontakte” are mostly people up to 30-35 years old. Right:
Chart 2. General distribution of Vkontakte users by age
Therefore, on the graph I will give two lines. The first is the distribution of the number of subscribers by age in absolute terms. The second is the “reference” distribution of subscribers, that is, how Habrahabr subscribers should be distributed by age if they were distributed in the same way as all Vkontakte users as a whole.
Graph 3. Absolute and “reference” values ​​of the distribution of Habr's subscribers by age
And so it was quite clear, I will give a graph of the percentage ratio of Habr's subscribers to the total number of all Vkontakte users of this age:
Graph 4.% of subscribers of the total number of Vkontakte users.
Quite a logical picture - Khabrhabr Vkontakte enjoys the greatest popularity among 20-25 year old people. The peak at 64 is due to the fact that the default when registering Vkontakte is set (or, at least, installed earlier) 1950 year of birth. But why such a jump is not on chart 2, I can not explain.
With age and gender, we figured out, we went to geography. For convenience, we divide the diagram into two different ranges (otherwise, nothing except the first four countries can be distinguished in the diagram). "Major" countries with the most subscribers:
Chart 5. Countries where Habrahabr has the most subscribers
And the "small" (confine ourselves to countries where there are at least 100 subscribers):
Chart 6. “Minor” countries of Habrahabr subscribers
The logical question is - firstly, where did I get that all those who indicated in their profile as their hometown New York really live in the USA, and not in some New Vasyuki? Yes, nothing. Do not check it in any way (at least in a simple way), so you have to take a word for it.
Secondly, the Habrahabr site is Russian-language. It is logical that it will be popular primarily among Russian users. But Vkontakte social network is popular mainly among Russian-speaking people. These are either residents of the ex-USSR, or emigrants, if we are talking about foreign countries. So let's apply the same trick as with age. Let's look at the percentage of subscribers to the total number of Vkontakte users from a particular country (I will call this ratio the “normalized percentage of subscribers” to save space and time. Maybe in SMM science there is some correct term for this concept - if there is, then tell me).
We look at the diagram and see curious things:
Graph 7. Percentage ratio of subscribers to the total number of Vkontakte users by country
Surprises Jamaica and Japan. I suspect anime and ahem are mixed in here ... let it be reggae music. Just look at the avatars of subscribers from Japan :)
In general, I will allow liberty and exclude them from our list as suspicious results.
Lyrical digression
By the way, VK has an interesting either bug, or feature, which is probably familiar to experienced VK API users, but for me it was a discovery and made me a little nervous.
If you just look for people from Japan among the subscribers, we get only 13 people:
However, if we also indicate the city, we will get a completely different picture:
Therefore, I studied information about countries based not on searching by group, but on data directly from user pages.
By removing these two wonderful countries, we get:
Chart 8. Standardized% of subscribers (excluding Japan and Jamaica)
The most interesting thing here is that neither Russia, nor Belarus, nor Ukraine even fall into the top three.
Russia shares its line with the Netherlands at all.
If Japan, Jamaica, or, for example, Honduras, we like to indicate how the country of residence is “purely funny,” then it is difficult for me to suspect that they are doing the same thing with Finland or Thailand. In this regard, I assume that, most likely, these are really subscribers living in these countries. Why, then, is Russia only in the middle of the list? I guess the whole thing in the following: who leaves abroad and at the same time actively uses social networks? I think that to a large extent highly qualified specialists in the field of IT, whom God himself ordered to be interested in Habrom. So it turns out that on average, Habr among Russian-speaking people living abroad is more popular than actually in Russia.
If you look at Europe (without ex-USSR countries), then the Habr's popularity map (countries with more than 100 subscribers are counted) will look something like this:
The following couple of paragraphs up to the section "Geography. Cities ”- my lyric reasoning on the topic of search, perhaps, is not very interesting and not very existing relationship. Who is not interested - just scroll further
Let's take a little bit of geography and economics. Let us turn our eyes to where the situation should be simpler, namely, to the countries of the ex-USSR. Here, the overwhelming majority of the Russian-speaking population, so it is possible to estimate Habr's popularity on VKontakte with a smaller share of mistrust.
Just looking at the countries of the former USSR is not very interesting, I want to understand why in one country or another the interest to Habr is higher. It would be logical to assume (hell, still could not resist the hypotheses, although he promised at the beginning) that the higher the level of development of information and communication technologies (ICT index) in the country, the more interest there is on the websites of such subjects (read to Habr ). ICT data are taken from here (no data are available for Tajikistan, Kyrgyzstan, and Turkmenistan).
At first the schedule:
Chart 9. The dependence of the normalized% of subscribers on the ICT index.
I do not even know what to say here. I planned that there would be a strong correlation and I, having played the role of Captain Obviousness, would simply voice this fact. However, the dependence is, frankly, not so hot (the Pearson correlation coefficient, for example, will be about 0.5). Ukraine stands out very much. The first thought was that in Ukraine all interest in Habra is concentrated only in a few cities or regions, which causes such a gap. However, if you look at the map of Ukraine, you will see that the normalized percentage is fairly evenly distributed:
I expected to see another picture in absolute numbers, but even here everything is more or less evenly:
So, either the ICT index is biased, or the people of Ukraine are interested in Habrom despite its low value, or it really has nothing to do with it.
By the way, if, for the sake of interest, to construct a graph similar to schedule 9 already for all countries (except Japan and Jamaica), we will get almost the same picture:
Chart 10. The dependence of the normalized% of subscribers on the ICT index (for all countries).
Let's go the beaten path. We will study the percentage of Habr's subscribers to the total number of Vkontakte users from this city. We will consider only those cities in which the number of subscribers is more than 10.
For a start, for the solemnity of the moment, look at the world map. Dots mark cities where the number of Habr's subscribers is more than 10. Some cities could not get on the map due to the discrepancy between the names of cities in the VK and the Google Chart API - I was too lazy to manually edit hundreds of names, so excuse me, the main thing is that the scope is clear!
Well, do not forget that this is all just the data from the social network, and not from the passport page. So, perhaps, everyone who pointed out Melbourne is actually living in my own Dolgoprudny.
By the way, I mentioned my hometown for good reason. Look at the top 30 Russian cities by the percentage of Habr's subscribers:
Chart 11. Normalized% of subscribers in Russian cities. TOP 30
I experienced such pride in my hometown a couple of years ago, when Dolgoprudny FC became the winner of the third division of the Russian Football Championship. I honestly did not customize. An accidental mistake or some peculiarity in the behavior of Vkontakte, which “pushes” more data in a search from your city, I can still admit, but this is unlikely.
I suspect that this figure is due to a large number of MIPT students (we’ll go further before).
If you look at the same diagram, but around the world, the picture will be more fun:
Chart 12. Standardized% of subscribers in cities of the world. TOP 30
I suspect that the data are not very reliable. But, at least, it is clear where many habravchane would dream to live :)
I, probably, already tired you with geography. Moving on.
First, we take the TOP-30 by the absolute number of subscribers (the university is hidden by many privacy settings, so do not be surprised by the small absolute numbers):
Chart 13. Number of subscribers by university
No surprises in my opinion. Now let's look, as always, at the normalized percentage of subscribers:
Chart 14. Normalized% of subscribers by university
Interestingly, in the first place Kazakhstani university. Of the Russian universities, as I assumed above, MIPT is the most haggard.
Another pair of diagrams that I could not do:
Chart 15. Political views
Chart 16. Alcohol and smoking attitudes
As we see, smoking is perceived worse than alcohol (however, this is a general contact, if not global, trend)
Chart 17. Marital status
What is interesting, in general, VK for people in a marriage is 1.5 times more than in the status of “single”. And the general distribution is different:
Chart 18. Marital status of VC users
The differences will even be statistically significant, but there is nothing surprising here - Habr's subscribers are mostly young people and, therefore, have not yet had time to get married legally.
So, let's summarize the small results and distribute the awards “for the lack of attention to Vkontakte”:
The most fascinated country is Belarus
The most crowded city is Palo Alto .
The most occupied city in Russia is Dolgoprudny.
The most enrolled university - MIPT .
Well, and you, a 24-year-old single graduate of Moscow State University, who has moderate political views and has no bad habits, I hope, it was at least a little interesting.
If my hands reach me, and it will be interesting to someone from habrovchan, in the next article we will analyze likes and reposts, I still have a lot of material.
Thanks for attention.
A year has passed since my previous article, and my grafomanskaya nature got the better of laziness and made me publish another small study.
The title of the article is, of course, slightly hypocritical (but it sounds beautiful). It would be more correct to call it this way: “Who is subscribed to Habrahabr Vkontakte”.
')
As you probably already understood, it was interesting for me to see who, from where and how Habr reads, likes Habr, repost Habr (however, the last two points in this article did not fit). But at the same time understand what cities, countries, universities, etc. the most "zahabrennye" in VK.
Who cares to see interesting charts, graphs and figures about this, please under the cat.
Warning
First, a couple of warning lines. I am not a professional sociologist, I am not a professional SMM researcher or web analyst. Some things I could overlook, some concepts can be called somehow unprofessional. For this, I ask you not to beat (at least in the face), but to point out in the comments on inaccuracies and comments. As far as possible, I will supplement and correct. Thank.
Also, I draw your attention to the fact that the sample under study is a public audience from the social network “Vkontakte”. This means that user data in it may periodically change, they may be incorrect or inaccurate. Therefore, when I say “Habr's readers are 146% of 91-year-old men from the Isle of Man”, this is not the ultimate truth. Just this is the information specified by users in the profiles.
One last thing: in this article I do not test any hypotheses. This article is rather a review, so there will be just graphs and diagrams, no verification of statistical significance, no regression models, no degrees of freedom and chi-squares.
UPD. 11/18/2014: so it turns out that on many graphs and maps appear the Crimea and its cities. By default, in Google GeoChart it is assigned to the territory of Ukraine, in VK - to Russia. Please do not look for any of my intent in attributing Crimea to this or that state, on all graphs and maps I simply left everything as default and unchanged. The active political position is good, but not in the form of sracha in the comments on Habré.
Okay, pretty chatting, let's get down to business.
Data and processing
Data downloaded using VK Api. They were pumped into a self-written database, where I processed and analyzed data. Initially, I wrote this database for the analysis of other groups and public Vkontakte, but when the work with them was finished, I could not resist and decided to analyze Habrahabr too.
However, this article is not devoted to the description of the technical aspects of the process of obtaining data from Vkontakte. There are enough examples of working with VK Api on Habré, and it’s not worth describing database design and development. If you have any questions - write.
Also, if there are any proposals to analyze some other parameters of users, about which I did not describe in the article or to subject any other public and groups to the same research, write in the comments.
The sample for the study is solid, that is, the entire audience of the public. The data were received in October 2014, so they may differ (most likely, slightly) from the current ones.
results
To simplify the narration, I will use the words “Habrahabr readers” or “Habrahabr subscribers” instead of the phrase “Habrahabr readers” (yes, in the context of this article, I’ll just say “subscribers”). Just remember that we are talking about the "contact" readers of Habr.
Floor
First, take a look at the floor of subscribers:
Chart 1. Distribution of subscribers by sex
The picture is quite logical. Habr is still a site for techies. Historically, there are more males among techies, and therefore there are more males among subscribers (despite the fact that in general on Vkontakte the ratio of men / women is about 53/47).
Age
Let's look at the number of readers by age (I took a range from 16 to 65).
But you will fairly note: so after all, active users of “Vkontakte” are mostly people up to 30-35 years old. Right:
Chart 2. General distribution of Vkontakte users by age
Therefore, on the graph I will give two lines. The first is the distribution of the number of subscribers by age in absolute terms. The second is the “reference” distribution of subscribers, that is, how Habrahabr subscribers should be distributed by age if they were distributed in the same way as all Vkontakte users as a whole.
Graph 3. Absolute and “reference” values ​​of the distribution of Habr's subscribers by age
And so it was quite clear, I will give a graph of the percentage ratio of Habr's subscribers to the total number of all Vkontakte users of this age:
Graph 4.% of subscribers of the total number of Vkontakte users.
Quite a logical picture - Khabrhabr Vkontakte enjoys the greatest popularity among 20-25 year old people. The peak at 64 is due to the fact that the default when registering Vkontakte is set (or, at least, installed earlier) 1950 year of birth. But why such a jump is not on chart 2, I can not explain.
Geography. Country
With age and gender, we figured out, we went to geography. For convenience, we divide the diagram into two different ranges (otherwise, nothing except the first four countries can be distinguished in the diagram). "Major" countries with the most subscribers:
Chart 5. Countries where Habrahabr has the most subscribers
And the "small" (confine ourselves to countries where there are at least 100 subscribers):
Chart 6. “Minor” countries of Habrahabr subscribers
The logical question is - firstly, where did I get that all those who indicated in their profile as their hometown New York really live in the USA, and not in some New Vasyuki? Yes, nothing. Do not check it in any way (at least in a simple way), so you have to take a word for it.
Secondly, the Habrahabr site is Russian-language. It is logical that it will be popular primarily among Russian users. But Vkontakte social network is popular mainly among Russian-speaking people. These are either residents of the ex-USSR, or emigrants, if we are talking about foreign countries. So let's apply the same trick as with age. Let's look at the percentage of subscribers to the total number of Vkontakte users from a particular country (I will call this ratio the “normalized percentage of subscribers” to save space and time. Maybe in SMM science there is some correct term for this concept - if there is, then tell me).
We look at the diagram and see curious things:
Graph 7. Percentage ratio of subscribers to the total number of Vkontakte users by country
Surprises Jamaica and Japan. I suspect anime and ahem are mixed in here ... let it be reggae music. Just look at the avatars of subscribers from Japan :)
In general, I will allow liberty and exclude them from our list as suspicious results.
Lyrical digression
By the way, VK has an interesting either bug, or feature, which is probably familiar to experienced VK API users, but for me it was a discovery and made me a little nervous.
If you just look for people from Japan among the subscribers, we get only 13 people:
However, if we also indicate the city, we will get a completely different picture:
Therefore, I studied information about countries based not on searching by group, but on data directly from user pages.
By removing these two wonderful countries, we get:
Chart 8. Standardized% of subscribers (excluding Japan and Jamaica)
The most interesting thing here is that neither Russia, nor Belarus, nor Ukraine even fall into the top three.
Russia shares its line with the Netherlands at all.
If Japan, Jamaica, or, for example, Honduras, we like to indicate how the country of residence is “purely funny,” then it is difficult for me to suspect that they are doing the same thing with Finland or Thailand. In this regard, I assume that, most likely, these are really subscribers living in these countries. Why, then, is Russia only in the middle of the list? I guess the whole thing in the following: who leaves abroad and at the same time actively uses social networks? I think that to a large extent highly qualified specialists in the field of IT, whom God himself ordered to be interested in Habrom. So it turns out that on average, Habr among Russian-speaking people living abroad is more popular than actually in Russia.
If you look at Europe (without ex-USSR countries), then the Habr's popularity map (countries with more than 100 subscribers are counted) will look something like this:
The following couple of paragraphs up to the section "Geography. Cities ”- my lyric reasoning on the topic of search, perhaps, is not very interesting and not very existing relationship. Who is not interested - just scroll further
Let's take a little bit of geography and economics. Let us turn our eyes to where the situation should be simpler, namely, to the countries of the ex-USSR. Here, the overwhelming majority of the Russian-speaking population, so it is possible to estimate Habr's popularity on VKontakte with a smaller share of mistrust.
Just looking at the countries of the former USSR is not very interesting, I want to understand why in one country or another the interest to Habr is higher. It would be logical to assume (hell, still could not resist the hypotheses, although he promised at the beginning) that the higher the level of development of information and communication technologies (ICT index) in the country, the more interest there is on the websites of such subjects (read to Habr ). ICT data are taken from here (no data are available for Tajikistan, Kyrgyzstan, and Turkmenistan).
At first the schedule:
Chart 9. The dependence of the normalized% of subscribers on the ICT index.
I do not even know what to say here. I planned that there would be a strong correlation and I, having played the role of Captain Obviousness, would simply voice this fact. However, the dependence is, frankly, not so hot (the Pearson correlation coefficient, for example, will be about 0.5). Ukraine stands out very much. The first thought was that in Ukraine all interest in Habra is concentrated only in a few cities or regions, which causes such a gap. However, if you look at the map of Ukraine, you will see that the normalized percentage is fairly evenly distributed:
I expected to see another picture in absolute numbers, but even here everything is more or less evenly:
So, either the ICT index is biased, or the people of Ukraine are interested in Habrom despite its low value, or it really has nothing to do with it.
By the way, if, for the sake of interest, to construct a graph similar to schedule 9 already for all countries (except Japan and Jamaica), we will get almost the same picture:
Chart 10. The dependence of the normalized% of subscribers on the ICT index (for all countries).
Geography. Cities
Let's go the beaten path. We will study the percentage of Habr's subscribers to the total number of Vkontakte users from this city. We will consider only those cities in which the number of subscribers is more than 10.
For a start, for the solemnity of the moment, look at the world map. Dots mark cities where the number of Habr's subscribers is more than 10. Some cities could not get on the map due to the discrepancy between the names of cities in the VK and the Google Chart API - I was too lazy to manually edit hundreds of names, so excuse me, the main thing is that the scope is clear!
Well, do not forget that this is all just the data from the social network, and not from the passport page. So, perhaps, everyone who pointed out Melbourne is actually living in my own Dolgoprudny.
By the way, I mentioned my hometown for good reason. Look at the top 30 Russian cities by the percentage of Habr's subscribers:
Chart 11. Normalized% of subscribers in Russian cities. TOP 30
I experienced such pride in my hometown a couple of years ago, when Dolgoprudny FC became the winner of the third division of the Russian Football Championship. I honestly did not customize. An accidental mistake or some peculiarity in the behavior of Vkontakte, which “pushes” more data in a search from your city, I can still admit, but this is unlikely.
I suspect that this figure is due to a large number of MIPT students (we’ll go further before).
If you look at the same diagram, but around the world, the picture will be more fun:
Chart 12. Standardized% of subscribers in cities of the world. TOP 30
I suspect that the data are not very reliable. But, at least, it is clear where many habravchane would dream to live :)
I, probably, already tired you with geography. Moving on.
Universities
First, we take the TOP-30 by the absolute number of subscribers (the university is hidden by many privacy settings, so do not be surprised by the small absolute numbers):
Chart 13. Number of subscribers by university
No surprises in my opinion. Now let's look, as always, at the normalized percentage of subscribers:
Chart 14. Normalized% of subscribers by university
Interestingly, in the first place Kazakhstani university. Of the Russian universities, as I assumed above, MIPT is the most haggard.
Political views, worldview, etc.
Another pair of diagrams that I could not do:
Chart 15. Political views
Chart 16. Alcohol and smoking attitudes
As we see, smoking is perceived worse than alcohol (however, this is a general contact, if not global, trend)
Chart 17. Marital status
What is interesting, in general, VK for people in a marriage is 1.5 times more than in the status of “single”. And the general distribution is different:
Chart 18. Marital status of VC users
The differences will even be statistically significant, but there is nothing surprising here - Habr's subscribers are mostly young people and, therefore, have not yet had time to get married legally.
Results
So, let's summarize the small results and distribute the awards “for the lack of attention to Vkontakte”:
The most fascinated country is Belarus
The most crowded city is Palo Alto .
The most occupied city in Russia is Dolgoprudny.
The most enrolled university - MIPT .
Well, and you, a 24-year-old single graduate of Moscow State University, who has moderate political views and has no bad habits, I hope, it was at least a little interesting.
If my hands reach me, and it will be interesting to someone from habrovchan, in the next article we will analyze likes and reposts, I still have a lot of material.
Thanks for attention.
Source: https://habr.com/ru/post/242951/
All Articles