跟读练习: Top 5 Insider Interview Questions Data Analysts Must Master Before Any Interview! - 通过YouTube学习英语口语
C1
I analyzed thousands of data analysts interview questions,
205 句
如果句子过短或过长，请点击 Edit 进行调整。
1
I analyzed thousands of data analysts interview questions,
2
and today we're going to go over the top five different types of questions
3
that I saw that you'll be expected to face in your next interview.
4
Number one, behavioral interview questions.
5
Behavioral interview questions are the most common type of question
6
that you're expected to see in almost any data analyst interview going forward.
7
Unlike data scientists that spend probably 70% of their time coding and doing IC work and maybe 30% talking to stakeholders,
8
data analyst work probably falls more in that 50-50 range,
9
where they're spending a lot more time actually interacting with people,
10
gathering requirements, creating charts, and collaborating with stakeholders.
11
Data analyst positions are generally pretty junior.
12
Another reason is that most people that apply for data analyst positions are generally more junior in their career.
13
So for a lot of new grads,
14
you're expected to see a lot of behavioral questions because they're trying to see how well you work with other people.
15
For example, some really common questions are,
16
how would you approach
17
x issue did you ever face y problem tell me about a situation where your time management went wrong
18
and how you fix it in answering these kinds of questions the easiest way to do
19
so is always to pick two
20
or three different stories where in your past experience you were
21
kind of depicted as a rock stone the best thing about
22
picking just a few different stories where you can rehearse them really well
23
and know them down to a t is that you can reapply these for all different kinds of question.
24
That same story that you use where you're showcasing your star project,
25
you can probably also talk about how you apply time management in there,
26
how you maybe you solve the conflict with another group member,
27
and it also describes how you work with other people.
28
Another way to approach this is to use a star framework,
29
which is situation, task, action, and result.
30
And I don't want to go too into it here.
31
You can always check out my other video on data science behavioral interview questions down below.
32
Now, I want to stop right here before going to the next question
33
because I want to add a caveat about how often the interview questions are going to show up for your interview.
34
One thing I noticed in my analysis of different company interviews is
35
that behavioral interview questions for data analyst interviews do come up more often depending on a variety of factors.
36
Namely, if the company that you're interviewing for is more non-tech,
37
so for example, something in finance like JP Morgan or airlines like American Airlines or even consulting like Deloitte,
38
A lot of the times they're going to emphasize behavioral interview questions more than the actual technical ones.
39
The more tech-focused companies like Facebook,
40
Amazon, or the fan companies are going to emphasize more technical questions compared to the behavioral ones.
41
And sometimes for a lot of these non-tech companies,
42
I've seen them ask only behavioral interview questions and they won't even ask you a single technical question at all.
43
The second thing to note is that the number of behavioral interview questions
44
that are asked is also correlated with generally less pay.
45
That's apparent because our next data analyst interview question type is on SQL.
46
Now SQL questions also range in difficulty and they also depend on what kind of data analyst interview you're going for.
47
So for example, pulling directly from Glassdoor and other sources,
48
you'll see companies like American Airlines will literally just ask you
49
on a scale of one to 10 how good your skill level is on SQL,
50
Python, and other technical skills.
51
They might also ask you an example of simple questions on SQL,
52
such as explaining what a CTE is,
53
or describing the difference between different types of joints.
54
These questions are meant to test your knowledge of SQL,
55
but they're not really expecting you to really test your effectiveness in writing SQL query.
56
Again, these questions show up for more non-tech companies,
57
whereas actual tech companies like Facebook and Amazon will give you case studies like these.
58
So for example, they'll give you two tables,
59
one with managers and one with employees,
60
and they'll ask you to write a query to find the manager with the biggest team side.
61
Now, these kinds of SQL questions are more like SQL case studies,
62
and companies like Amazon for their data analyst interview will even go a step further with their technical screen on SQL questions.
63
They'll ask you first a question like the manager team size is one,
64
and then they'll give you harder and harder questions.
65
You're given the same two tables.
66
Now write a query to find all the employees that join before the manager.
67
For the Amazon data analyst interview,
68
they'll give you five concurrent questions,
69
each one a little bit harder than the next.
70
Each candidate is then benchmarked on how well they do.
71
So if you can solve all five questions within the amount of time for the technical interview,
72
then they'll generally pass that interview.
73
But over time, if they see that more candidates are passing the technical screen pretty easily,
74
then they'll make those questions slightly harder.
75
Because at the end of the day,
76
Amazon is trying to filter out around 50% of the candidates at every step of the interview.
77
If you want to check out more SQL interview questions,
78
check out my video below where I've conducted mock interviews with
79
multiple different kinds of data scientists at top tech companies on SQL.
80
The next common question that you're going to see is the analytics case study question.
81
Now analytics case study questions can comprise of various levels of difficulty.
82
The easiest ones are generally the most straightforward where they're asking you a theoretical case.
83
So for example, what metrics would you use to measure the success of a product?
84
Another one would be something like,
85
we've seen a drop in conversion rate over the last week,
86
how would you investigate what would happen?
87
Now while these are more theoretical in nature,
88
sometimes you'll actually have to go into the actual data set that the interviewer provide and analyze the actual data.
89
And I really like these questions myself as an interviewer
90
because these give me a two-for-one kind of case study where I can figure out,
91
one, how good their analytical reasoning skills are,
92
and two, exactly how good their actual SQL or technical skills are as well.
93
So let's take an example question.
94
So you're given a table that represents search results from all searches on Facebook.
95
The query column is a search term and the position column represents each position their search result came in.
96
And lastly, the rating column represents human of a search result from one to five,
97
but five is high relevance and one is low.
98
So in this example, we have to solve two things.
99
One is we actually have to create the metric on our own,
100
given the problem requirements.
101
And the second is that we also have to write the query.
102
So let's try to tackle the first part.
103
How do we construct a metric that actually applies a good rating for every search result?
104
Well, let's look at the problem.
105
For example, if the first result is rated at a five and the last result is rated at a one, that's good.
106
But it'd be better if the first result is rated five and the last result is also rated five.
107
The worst example scenario would be if the first result is rated as a one and the last result is a file.
108
So looking at this, I guess the actual answer
109
that we should do is we should inverse the position as a weighted factor.
110
In this case, we would apply one over the position as a weighted score.
111
Now, no matter what the overall rating is,
112
we have a way to weigh the position into the formula.
113
Let's take another example.
114
So let's say that you're given event data from a social networking site like Facebook.
115
A product manager is interested in understanding the average number of sessions that occur every day.
116
However, the company has not technically defined what a session is yet.
117
Given a data set of events,
118
how would you define what a raw session looks like for a company like Facebook?
119
Then write a query to return the average number of sessions per day.
120
So again, from our example here, we're doing two things.
121
One is we're actually defining the metric at large,
122
and then two is that we're actually going through and writing a query for this example schema.
123
The next common interview question that we see on the interview is Python.
124
So Python questions comprised of two different types.
125
One is data manipulation that involves pandas.
126
So that means basically taking a SQL interview question
127
and transposing it into more of a pandas interview question because they both deal with data manipulation.
128
The second one is more around general kind of string manipulation,
129
data manipulation that you can do with just using other data structures in Python.
130
The reason why Python is so popular now is
131
that data analysts have slowly transitioned away from R and S-A-S as kind of the main tools for data analysis,
132
given that Python now has much more modularity when it comes to working with different kinds of data analytics tools,
133
dashboards like Streamlit, and also being able to implement anything
134
that you need more in a script in production easier than scaling out languages like R.
135
The other thing to note is that most Python interview questions are going to be fairly easy.
136
You won't expect anything from leak code in your data analyst interview.
137
At the end of the day,
138
the most that data analysts will work on in pipeline is probably writing scripts and not really scaling anything in production.
139
Example question that was asked recently at Deloitte would be,
140
given a list of integers,
141
identify all of the duplicate values in the list.
142
So pretty standard.
143
A solution here would be to create a dictionary and add all of the integers as the keys,
144
and then the counts of the actual duplicates would be anything greater than one.
145
And those would be the values that we would return.
146
Another really common Python data analyst interview question we've been seeing recently is one around string shift.
147
So given two strings, A and B,
148
write a function can shift to return whether
149
or not A can be shifted to the number of places to get to B.
150
The last question that we see on data analyst interviews are around statistics and A-B testing type interview questions.
151
Baseline of a lot of data analytics work is obviously fundamentally based in stats.
152
Specifically, there's a couple of concepts that we wanna talk about here, right?
153
Cause there's the causal inference part of statistics and then there's also the A-B testing part,
154
which is more experimentation.
155
At the base of both of those is just general statistics questions
156
that you might perceive if you're interviewing for an entry-level data analyst.
157
The company might ask you,
158
what are Z tests and T tests?
159
What's the difference between them and when should we use either of them?
160
A t-test is a statistical test
161
that uses the t family of distributions to compare two means to see if they're significantly different from each other.
162
And they're typically conducted when the population standard deviation is unknown.
163
So generally these kinds of questions are gonna be more definition-based.
164
You're just gonna have to remember these in general,
165
especially if you're expected to have a more broader and fundamental statistical analysis background.
166
More practical case study is one where the solution the case study is actually a type of formula
167
that you might have to know
168
so for example the capital approval rates have gone down for
169
our overall approval rate let's say last week is 85
170
and the approval rate went down 82 this week
171
which is a statistically significant reduction the first analysis shows
172
that all the approval rates stayed flat or increased over time
173
when we have products product one 84 85 percent week over week product 277 to 77 product 381 82
174
and product 488 to 88 what could be the cause of the decrease now
175
if you know you know
176
and for example this solution the actual answer is the simpson's paradox the simpson's paradox is a phenomenon
177
when basically a trend shows in several groups either disappears
178
or is reversed when combining the data and that's
179
because the size of some of these groups is maybe larger than the size of other groups
180
which causes conflation when we merge all the data together wow this is a lot of information
181
and it's a lot to How do I actually prepare for my next interview?
182
I would then highly recommend InterviewPro,
183
which is a company that I founded five years ago.
184
We are the number one interview prep platform for data analytics,
185
and we have hundreds of data analysts,
186
interview questions, and course material on data analytics as well.
187
Specifically, we have a learning path that you can take
188
that has over 10 plus hours of course content that guides you through causal inference,
189
inference, data analytics case studies,
190
and different types of interview questions like statistics, SQL, Python, and more.
191
But don't just take my word for it.
192
I know it sounds great and stuff.
193
Let's look at Siobhan, who actually landed a job using the interview query.
194
He sent me an email that said, Hi, Jay.
195
I actually use your videos and platform to help me prepare for my interview.
196
And I got a job as a data analyst and product analytics at Autodesk.
197
Before my interviews, I would watch your videos and go over any questions they may ask.
198
Thank you for your work on the platform.
199
I had a friend interviewing for data jobs and he's an engineer,
200
so I recommended you get your platform.
201
We're so happy about our success stories.
202
And if you're looking for facing your next interview,
203
obviously check out Interview Query.
204
You can use the link below.
205
Thanks for watching everyone.
下载应用
AI 为你说出的每个句子打分
TRENDING
热门
关于本课
在本课程中，学习者将通过分析数据分析师面试中常见的行为面试问题，提升自己的英语口语能力。我们将主要关注如何有效回答这些行为面试问题，以展示个人的沟通能力和团队合作精神。通过这项练习，您将能够了解面试官希望看到哪些能力，同时提升您的英语发音和表达能力，助力雅思口语练习。
关键词汇和短语
- 行为面试问题 (behavioral interview questions)
- 团队合作 (team collaboration)
- 时间管理 (time management)
- 技术性问题 (technical questions)
- SQL (数据库查询语言)
- 数据分析师 (data analyst)
- 情况-任务-行动-结果框架 (STAR framework)
- 沟通能力 (communication skills)
练习建议
在本课程中，建议您进行shadow speech练习。可以选择视频中的片段，以模仿讲者的语调、语速和停顿。特别是在面试的问题中，强调回答的shadowspeak会让您的表达更具信服力。您可以尝试以下步骤来提高您的英语发音：
- 选择一段适合您的水平的节选。
- 在听的过程中，注意发音的细节和语调的变化。
- 暂停视频并尝试重复讲者的每一句话，确保模仿其语速和情感表达。
- 多次重复每个句子，提高您的流利度。
- 可以录下自己的声音，回放以确认发音的准确性和自信程度。
通过这些方法，您不仅能提高您的英语发音，还能练习如何在面试中自信地回应问题，进而在实际应用中形成良好的口语表达习惯。练习shadow speak也是增加语感的绝佳方式，帮助您在雅思口语练习中脱颖而出。
什么是跟读法？
跟读法 (Shadowing) 是一种有科学依据的语言学习技巧，最初开发用于专业口译员的培训，并由多语言者Alexander Arguelles博士普及。这个方法简单而强大：您在听英语母语原声的同时立即大声重复——就像是一个延迟1-2秒紧跟说话者的影子。与被动听力或语法练习不同，跟读法强迫您的大脑和口腔肌肉同时处理并模仿真实的讲话模式。研究表明它能显着提高发音准确性，语调，节奏，连读，听力理解和口语流利度——使其成为雅思口语备考和真实英语交流最有效的方法之一。