Use full-stack developer calculations

Are fewer questions sometimes more?

Also this year the popular open source community freeCodeCamp carried out its New Coder Survey. freeCodeCamp has made a name for itself through donation-financed tutorials as well as camps and meetings for intensive learning of programming languages. The survey is intended to statistically record various parameters such as level of education, career aspirations, acceptance of tutorials and camps, as well as origin and life situation of young programmers. Some interesting facts, but also a few trivialities, were brought to light. But first things first, about the method and the evaluation of the study. Who are the so-called new coders?

Purpose of the study

The second time the study was carried out, the target group of the study was the New Coders: This means beginners who have been a maximum of five years program or learn to program. This time 50 questions were asked (in 2016 there were 48). At the same time, not everything had to be answered; surveys with many blank spaces were also included in the database.

All data was recorded anonymously in a CSV file and then made available free of charge on a GitHub repository under an Open Data Common License. The authors' calculation has already proven its worth: Last year, the liberal approach led to many external data scientists and data engineers working on the mountain of data and publishing dozens of evaluations.

Demographic background

Compared to the previous year, the number of participants has increased by a full third, from 15,000 to 20,000. That is good news for the study's authors, but it has brought about noticeable changes. Initially, the proportion of women among the beginners surveyed fell from 21 to 19 percent, which is still twice as many as in other studies. The participants are on average 28 years old and have been coding for 21 months. So the positive thing is that you can start learning to code in your late twenties or even later.

The relative number of Americans has decreased from 38 to 44 percent, but it has increased from 6,200 to 6,900. The second largest group comes from India with ten percent (nine percent in the previous year), followed by the British with five percent (four percent in 2016). At the same time, the proportion of English as the language used at home has decreased from 54 to 46 percent. Around 24 percent belong to an ethnic minority in their country. The bottom line is that freeCodeCamp can boast of having a growing and diverse customer base.

Level of education and other social factors

8 percent have served as conscripts or in other forms in the military. Most new coders already have a university degree: 54 percent of those surveyed have at least a bachelor's degree (58 percent in the previous year). Many are in IT courses, but social scientists and economists are also learning to program. To do this, 23 percent had to take out a student loan, and eleven percent had a mortgage on their property.

The fact remains that IT does not take place in the village: 43 percent of the bereaved live in large cities with over one million inhabitants (2016: 41 percent), 23 percent in small towns with fewer than 100,000 inhabitants and 35 percent in medium-sized cities (36 Percent). Urbanity brings with it the corresponding relationship status: two out of three respondents are single. However, 15 percent have children to look after (2016: 18 percent) and six percent support other family members such as parents and grandparents.

Overqualified, dreams of medium-sized businesses

The new survey provides a more detailed picture of the aspired developer professions because, in contrast to the previous year, several answer options can be selected. 59 percent of those surveyed want to work as full-stack developers, 50 percent in the front-end area and 40 percent in the back-end area. On the other hand, 33 percent would like to become mobile developers, 24 percent data scientists and 21 percent UX designers. For the first time, Game Designer is in its own category, which was also ticked by 23 percent. The prospering field of activity of security, which 19 percent are aiming for, has also been added.

Corporations are not a dream destination

Google and Co.? Are you kidding me? Are you serious when you say that! Large corporations are only 13 percent of the employers of choice. 22 percent would prefer to work for a medium-sized company, while 21 percent would like to start their own, 19 percent dream of becoming self-employed and 16 percent would prefer to start at a start-up. Even more than 30 years after Bill Gates and Steve Jobs, the ideal of many programmers is still somewhere between home office and garage.

Another picture shows the actual employment situation of the respondents: around two thirds are employed in some form. 52 percent are actually in an IT job (2016: 50 percent). But only 25 percent have a real developer job (2016: 28 percent). Ultimately, 44 percent of those surveyed think that they are overqualified for their current job (42 percent). 74 percent are ready to move for a new job (77 percent). The commuter status was also asked and 38 percent actually drive at least one hour to work. Combined with the debt, it can be said that many would-be programmers go to great lengths to achieve their dream job.

Learn to program - on all channels

The most popular resources for coding are, besides freeCodeCamp with 72 percent, above all Stack Overflow with 62 percent and W3Schools with 54 percent. Codecademy comes to 53 percent and Mozilla’s Developer Network to 35 percent, Udemy to 29 percent and CSS Tricks to 26 percent. But the survey also focused on media other than blogs and portals.

Among the podcasts, Code Newbie are the most popular with 24 percent and JavaScript Jabber 27 percent. Overall, the number of podcast listeners among those surveyed has risen sharply, from 26 percent in the previous year to 37 percent. Three quarters of respondents also watch videos. When asked about the most popular YouTube channels, freeCodeCamp received the most votes with 46 percent, followed by Google Developers with 27 percent and MIT Open Courseware with 26 percent.

Another aim of the survey was to investigate the role of boot camps and other programming seminars. Less than half have personally attended a boot camp and only six percent have been to an intensive camp. However, 82 percent of bootcamp participants were satisfied and would recommend the experience to others. However, 33 percent had to borrow money to attend. This has often paid off, as 58 percent of bootcamp participants now have a full-time job as a developer.

The bottom line and the limits of transparency

The conclusion of the authors after an initial evaluation is that the independent allocation of resources for learning programming is extremely important. It remains unspoken that the study should scientifically substantiate that code camps are a sensible investment. This does not screen your own customers, but of course also advertises your own offers, and if this result is confirmed by other analysts and carried out into the world, all the better for freeCodeCamp.

As well as the transparent open source study may be meant well - neutrality and meaningfulness must be critically questioned if all the questions asked have been answered with "freeCodeCamp is best". After all, open source should not only create a semblance of neutrality, but actually create it.

The problem can also be seen in the change in the social structure of the participants: If the participants are on average a year older and have now been coding for 21 months instead of eleven months, there is no need for a certificate in statistics to interpret that this is predominantly a very similar regular audience was surveyed again. That may of course make planning your own offer easier for them. At the same time, however, nobody should expect too many new findings compared to last year.

Less questions are more

In addition, the 2017 survey wasted many a question on private and irrelevant issues. It is an open secret that fewer questions result in more fully completed questionnaires. One can argue about which topic is important, but the technical and pedagogical aspects should remain in the foreground and be asked as concisely as possible. The question of whether the respondents were in the military only becomes valuable if it is also questioned whether it was used to finance studies or code camps.

So instead of preparing a similar mess of data every year, other options should be considered in the meantime: Elaborate qualitative interviews could provide precise conclusions about how exactly a young programmer went through his or her life, what the greatest obstacles were and where it should go. Or also why freeCodeCamp has a high proportion of women. So there is still a lot of potential in this matter and in the future it should also be considered whether, open source or not, an independent portal might not deliver more neutral results.