Before you continue reading I must state that one should be suspicious of the predominantly positive ratings about Lambda School on this platform because it is biased towards individuals asked to complete reviews after finding employment. People who have bad experiences may not even bother finding review sites like these and posting on them. Also, I can assure you that the signals of the internal level of satisfaction with Lambda from students (based on comments by students within the ...
Before you continue reading I must state that one should be suspicious of the predominantly positive ratings about Lambda School on this platform because it is biased towards individuals asked to complete reviews after finding employment. People who have bad experiences may not even bother finding review sites like these and posting on them. Also, I can assure you that the signals of the internal level of satisfaction with Lambda from students (based on comments by students within the platform on Slack suggest a significantly lower true approval rating).
About a year ago a Lambda graduate wrote a
reddit post that criticized Lambda School (henceforth "Lambda") for:
- cohort size
- inflating positive PR by encouraging students to say good things about LS on social media platforms
- not having a proven track record (due to the Lambda's relative youth compared to, e.g., App Academy)
This post was followed by a comment that said the real problem was:
- Negative students stirring up trouble
- Over-reliance on a "teaching you how to learn" Motto
This is by no means the only reddit post that says less than pleasant things about Lambda, but it's likely the top one if one Googles something like "should I join lambda school?". I won't agree or disagree with any of the statements in that linked post, but I think it forms a useful backdrop for my critique of the
Lambda School Data Science program.
The purpose of my post is to offer a review of Lambda School that is 1)
specific to the Data Science program and 2) not constricted within the heavily-curated official Lambda subreddit. When I wanted to join the Lambda DS program, there were very few useful reviews (most were web-dev-centric), so I hope I can help others who are researching the program.---------------------------- Last year around this time Udacity did a major overhaul to its pricing structure that made it no longer an affordable option for me to use to pursue a career in data science (DS). I went on a google search to find cheaper alternatives and I found a reddit post about the Udacity price hike in which one of the commenters said something to the effect that Udacity's new pricing would make schools like Lambda more popular for career training. I Googled "Lambda School" and was sold on the
length of the program, the
topics covered, and the
income share agreement (ISA) that would allow me to defer tuition costs until after I found a job paying $50,000 or more in a relevant tech field.I considered the length of the program (7-9 months) good because I personally would be more convinced about someone saying they obtained a certain skill set from a 7-9 month program than from a 3-month program (which is a common time period for many bootcamps). I liked that the program would spend a lot of time on statistics (4 weeks) instead of only introducing it as part of the EDA on an otherwise Machine-learning focused curriculum. I don't have any links to post you may see a lot of complaining online about DS education not being stats-focused. I appreciated the attempt at spending more time on the subject. Lastly, the ISA meant I went from allocating a budget for a lower-cost version of Udacity to a free-for-now (deferred cost) program, which allowed me to use that money to pay off my apartment lease for one year. This was advantageous since fulltime Lambda programs are time intensive (11 to 7 EST, in theory, but 11-9pm EST in reality)--so not not much time to earn an income (few jobs would be down for working around a schedule like that anyway, FME).
Good things about the Lambda Data Science Program:
-
The Slack-based Learning Platform: Lambda school has a student dashboard webpage that may have been made to be the source of knowledge/information, but everything really goes down in the Slack app. While the instructor is teaching via Zoom, you can ask questions on Slack that can be answered by the instructor (more on this later) or answered by other students. Questions and comments are then visible and can be pinned to the Slack Channel for future reference. For anyone who has experienced the asynchronous "discussion board" layout of many online learning institutions (I have), knows the value of a synchronous learning environment. A good instructor will make excellent use of the to ability communicate with students while teaching by asking and answering questions, creating polls (including anonymous ones), and interacting positively by liking/adding emojis to certain comments. The emoji thing may seem trivial but the more interaction the better and an approval signals to others what is the "best answer".
-
The chance to interact with people who are in different stages oflearningandemployment. A Lambda Student will be assigned to a group with is led by a team lead (TL) who is former Lambda student who is still on the job hunt. They are not DS experts yet, but they have the potential to provide useful perspectives and even learn more themselves (as a lot can be learned by teaching). As a lambda DS student, you will also have access to Slack channels where students in other cohorts (ahead of and behind your own cohort) discuss various DS-related topics. Students who secure employment can remain active in communal Slack channels after securing employment.
Also, you can actually talk to real Data Scientist during weekly "Brown Bag" Lunch presentations from guest experts. I personally never had time to attend a "Brown bag" session (and I don't see how a serious student could have such time, but the brown bags do exist).I suppose all of 2) still ties into the fact that Slack is used as the main hub for Lambda school and so one can avail of all the features Slack offers.3.
The openness of administration to conversations about improvements to Lambda. There is a particular Slack Channel called "announcements" in which a new change/service is announced, and students can respond to it in a related channel called "announcement threads". An outsider would be shocked at the kinds of conversations that happen here in which students to varying levels of zeal, express how they feel about an actual or potential change. Administrators are exposed to a level of criticism that one can't help sympathize with but a lot is learned in the process. Also students (including grad) can and will offer criticism about specifically the DS curriculum in relevant slack channels. They are often a fun read after getting through a day of hard work.So far so good, right? Ready to enroll? Not so fast...
Bad things about the Lambda Data Science Program:One might note that the benefits of Slack are more or less independent of Lambda School's educational quality. Meaning, putting a bad driver in a Lamborghini doesn't make them a good driver, but they will have more options and might be more motivated to learn to become a good driver (why let such a good vehicle go to waste?)
-
Admissions: It becomes apparent very quickly that not all students completed or understood the assignments that are supposed to be completed prior to being accepted into a cohort. I won't delve too heavily into this subject because such criticism about Lambda are echoed elsewhere. I do want to clarify, however, that my issue with whether or not a student was "smart" enough to be a part of the program. Instead, my issue is with, how at the very beginning, they did not demonstrate level of effort needed to be part of an intensive educational program. I don't care if they have to complete the precourse work five times, make them do it until they pass, and it can be certain that someone who constantly failed the precourse work, but stuck through it
-
Poorly evaluating students who in the program: Once students are in the program, they still are susceptible to being passed through even if their understanding is low or work is mediocre. The problems with this is the same as those mentioned in 1)
- Confusing "Expert" with "Teacher"
The Lambda DS program, at the time of this post, has an energetic, interactive teacher, Ryan A., who teaches Unit 1. He makes full use of the synchronous learning environment that slack provides. Ryan
stays on topic (or keeps his tangents short), constantly asks questions to check for understanding, creates comprehensive end-of-week exams that are sufficiently more difficult than the daily assignments (so you're not just copying and pasting and editing), and consistently makes himself available to video conference Q&A sessions during the time allotted for completing the daily assignment. It's true that he's a data scientist, but through either talent or effort has confirmed that he is a teacher as well.I won't bother calling out any names about the teachers that proceed Ryan A., but I can promise you that the learning experience after Ryan is a downhill tumble stemming mostly from the fact that there doesn't seem to be any standardized protocols or metrics for judging how well a teacher has performed except for feedback forms on which, one a week, ask you to rate your instructor and leave a comment if you wish.You will see a lot talk about how the CS section of Lambda school kills off a lot of students...many students attribute the problem to the curriculum. I can't speak for the other programs but
the CS section the DS program (unit 3) doesn't have a curriculum problem, it has a teacher problem. I have no doubt the teacher in question has a CV that qualifies him for the role of "data scientist", but his level of interaction with students, ability to seek out gaps in understanding, ability to organize his allotted instructional time, eagerness to interact with students after class, ability to multitask (by looking at both slack and text editor, ability to live code instead of editing pre-coded material, ability to explain his steps as he does, ability handle criticism, are all LOW. This is the simple truth.The main point, even if it is against common belief, is that it is better to have a person with mediocre domain expertise who is a good teacher than an "expert" who is a poor teacher if the chief goal of an institution is to provide others with an education (think about a manager who "knows everything" but is despised...how effective could she/he be?) . Otherwise, you might find leaning too heavily upon pedagogical philosophies of "learning how to learn" when signs point to a lack of understanding among the students/customers who are banking their future incomes on the quality of your education.----------------------
Today,
I quit the DS program even though I had passed all previous assignments and could easily be in the 85% percentile of my cohort in terms of understanding the material. I personally have no patience for a consistently poor learning experience and unfortunately Lambda has its DS program setup such that the same instructors teach the same units. So, at the moment, students will continue to enter the program with a false sense of how their learning experience will continue due to either a coincidental or clever decision to make the best teacher the teacher for unit 1. I say "clever" because if you don't withdraw from Lambda within 4 weeks, you will still owe part of your future income to the company (and each unit is 4 weeks long).I think the
potential for Lambda's DS program to garner a reputation that catches the eyes of employers is high. I speak, nevertheless, only of potential because at the moment there are some serious pedagogical concerns that need to be addressed. I might suggest something like peer training (where different teachers show how they would teach the same lesson, for example). Ideally, there would be 360 feedback (teacher to student, student to teacher, and teacher to teacher) but any attempt to be more vigilant of instructional quality would make the program not worth dropping out of. Until then, there current platform is easily replicable by those who want to also get into the DS Bootcamp industry (or those who are already in it). Recipe: Slack, Zoom, small group of instructors, and a team of administrators, and some venture capitalists. Emphasis on the Slack and Zoom.I hope others find this helpful and I trust Lambda will eventually improve its DS program. I like the CEO and the leader of the DS program (Aaron). They seem to have their hearts in the right place, but education first, guys.