Week 1: Effective questions

Introduction to problem-solving and effective questioning in data analysis

Data analysis is the process of collecting, cleaning, and interpreting data to extract insights. It is a critical skill for many different professions, including business, science, and government.

Problem-solving and effective questioning are essential skills for data analysts. By being able to identify and define problems clearly, and by asking the right questions, data analysts can develop and implement effective solutions.

Problem-solving in data analysis

The first step in problem-solving in data analysis is to identify and define the problem clearly. What are the specific questions that you are trying to answer? What data do you need to collect? What are the constraints on your analysis?

Once you have a clear understanding of the problem, you can start to develop a solution. This may involve collecting new data, cleaning and preparing existing data, and using statistical methods to analyze the data.

It is important to be iterative in your problem-solving process. As you learn more about the data, you may need to refine your questions or change your approach.

Effective questioning in data analysis

Effective questioning is essential for data analysts to be able to extract insights from data. By asking the right questions, data analysts can identify patterns and trends that would otherwise be overlooked.

Some examples of effective questions that data analysts might ask include:

• What are the most important factors that influence customer churn?
• What are the most effective marketing channels for reaching new customers?
• What are the key performance indicators (KPIs) that we should be tracking to measure our success?
• What are the risks associated with our current business model?

By asking effective questions, data analysts can develop a deeper understanding of the data and identify areas for improvement.

Conclusion

Problem-solving and effective questioning are essential skills for data analysts. By being able to identify and define problems clearly, and by asking the right questions, data analysts can develop and implement effective solutions, and extract insights from data to improve decision-making.

Here are some tips for developing your problem-solving and questioning skills in data analysis:

• Be clear about your goals. What do you want to achieve with your analysis?
• Understand your data. What are the different types of data that you have? What are the strengths and limitations of each type of data?
• Use appropriate statistical methods. There are many different statistical methods available, so it is important to choose the right method for your analysis.
• Be iterative. Don’t be afraid to refine your questions or change your approach as you learn more about the data.
• Ask for help from others. If you are struggling to solve a problem, don’t be afraid to ask for help from a mentor or colleague.

By following these tips, you can develop your problem-solving and questioning skills and become a more effective data analyst.

Welcome to the second course in
the Google Data Analytics certificate. If you completed Course One,
we met briefly at the beginning, but for
those of you who are just joining us, my name is Ximena, and
I’m a Google Finance data analyst. I think it’s really wonderful
that you’re here with me learning about the fascinating
field of data analytics. Learning and education have
always been very important to me. When I was young, my mom always said,
“I can’t leave you an inheritance, but I can give you an education that opens doors.”
That always pushed me to keep learning, and that education gave me the confidence
to apply for my job at Google. Now I get to do really
meaningful work every day. Just recently I worked as an analyst
on a team called Verily Life Sciences. We were helping to get life-saving medical
supplies to those who need it most. To do this, we forecasted what health
care professionals would need on hand and then shared that information with networks. The information that my
team provided helped make data driven decisions that
actually saved lives. I’m thrilled to be your instructor for
this course. We’re going to talk about
the difference between effective and ineffective questions and
learn how to ask great questions that lead to insights that can help
you solve business problems. You will discover that effective questions
help you to make the most of all the data analysis phases. You may
remember that these phases include ask, prepare, process, analyze, share, and act. In the ask step,
we define the problem we’re solving and make sure that we fully understand
stakeholder expectations. This will help keep you
focused on the actual problem, which leads to more successful outcomes. So we’ll begin this course by
talking about problem solving and some of the common types of business
problems that data analysts help solve. And because this course focuses on
the ask phase, you’ll learn how to craft effective questions that help you collect
the right data to solve those problems. Next, we’ll talk about the many
different types of data. You’ll learn how and
when each is the most useful. You’ll also get a chance to explore
spreadsheets further and discover how they can help make your data
analysis even more effective. And then we’ll start learning
about structured thinking. Structured thinking is the process of
recognizing the current problem or situation, organizing
available information, revealing gaps and opportunities,
and identifying the options. In this process, you address a vague, complex
problem by breaking it down into smaller steps, and then those steps lead you to
a logical solution. We’ll work together to be sure you fully understand how to
use structured thinking and data analysis. Finally, we’ll learn some proven strategies
for communicating with others effectively. I can’t wait to share more about my
passion for data analytics with you, so let’s get started.

Data analytics is the process of collecting, cleaning, analyzing, and interpreting data to extract meaningful insights. Data in action is when data is used to make decisions or take action.

There are many ways that data can be used in action. Here are a few examples:

• Business intelligence: Businesses use data analytics to make better decisions about their operations. For example, a retail store might use data analytics to determine which products are most popular and where to place them in the store.
• Fraud detection: Financial institutions use data analytics to detect fraud. For example, a bank might use data analytics to identify suspicious transactions.
• Customer relationship management (CRM): Companies use data analytics to manage their relationships with customers. For example, a company might use data analytics to identify customers who are likely to churn.
• Healthcare: The healthcare industry uses data analytics to improve patient care. For example, a hospital might use data analytics to identify patients who are at risk for developing certain diseases.
• Marketing: Marketers use data analytics to target their advertising campaigns more effectively. For example, a company might use data analytics to identify people who are likely to be interested in their products.

These are just a few examples of how data can be used in action. As the amount of data available continues to grow, the possibilities for using data in action will only increase.

Here are some tips for using data in action:

• Start with a clear goal in mind. What do you want to achieve by using data? Once you know your goal, you can start to collect the right data and analyze it in a way that will help you achieve your goal.
• Use the right tools. There are many different data analytics tools available. Choose the tools that are right for your needs and that will help you get the most out of your data.
• Be patient. Data analytics can be a complex process. Don’t expect to get immediate results. Be patient and persistent, and you will eventually see the benefits of using data in action.

Marketing analytics is the process of measuring, analyzing, and managing a company’s marketing strategy and budget. Often, this involves identifying the company’s target audience.
The target audience includes which people?

The people the company is trying to reach

The target audience is the people the company is trying to reach.

In this video, I’m going to share an interesting data
analytics case study, it will illustrate how problem solving relates to each phase of the data analysis
process and shed some light on how these phases
work in the real world. It’s about a small
business that used data to solve a unique
problem it was facing. The business is called
Anywhere Gaming Repair. It’s a service provider
that comes to you to fix your broken video game
systems or accessories. The owner wanted to
expand his business. He knew advertising is a proven way to get
more customers, but he wasn’t sure
where to start. There are all kinds of different
advertising strategies, including print, billboards,
TV commercials, public transportation,
podcasts, and radio. One of the key things to
think about when choosing an advertising method is
your target audience, in other words, the specific people you’re
trying to reach. For example, if a medical
equipment manufacturer wanted to reach doctors, placing an ad in a health magazine would
be a smart choice. Or if a catering company
wanted to find new cooks, it might advertise using a poster at a bus stop near
a cooking school. Both of these are
great ways to get your ad seen by your
target audience. The second thing to think
about is your budget and how much the different
advertising methods will cost. For instance, a TV ad is likely to be more expensive
than a radio ad. A large billboard
will probably cost more than a small poster
on the back of a city bus. The business owner
asked a data analyst, Maria, to make a recommendation. She started with
the first step in the data analysis process, Ask. Maria began by defining the problem that
needed to be solved. To do this, she first had to zoom out and look at the whole
situation in context. That way she could be sure
that she was focusing on the real problem and
not just its symptoms. This leads us to
another important part of the problem solving process, collaborating with stakeholders and understanding their needs. For Anywhere Gaming
Repair, stakeholders included the owner, the vice president
of communications, and the director of
marketing and finance. Working together, Maria and the stakeholders
agreed on the problem, not knowing their target audience’s preferred type of advertising. Next step was the prepare phase, where Maria collected data for the upcoming analysis process. But first, she needed to better understand the company’s
target audience, people with video game systems. After that, Maria collected data on the different
advertising methods. This way, she would be able
to determine which was the most popular one with the
company’s target audience. Then she moved on to
the process step. Here Maria cleaned
the data to eliminate any errors or inaccuracies that could get in the
way of the result. As we’ve learned,
when you clean data, you transform it into
a more useful format, create more complete information
and remove outliers. Then it was time to analyze. In this step, Maria wanted
to find out two things. First, who’s most likely to
own a video gaming system? Second, where are these people most likely to see
an advertisement? Maria, first discovered that
people between the ages of 18 and 34 are most likely to make video
game related purchases. She could confirm that Anywhere Gaming Repair’s
target audience was people 18-34 years old. This was who they should
be trying to reach. With this in mind, Maria then learned that
both TV commercials and podcasts are very popular with people in the
target audience. Because Maria knew
Anywhere Gaming Repair had a limited budget and understanding the high
cost of TV commercials, her recommendation
was to advertise in podcasts because they are
more cost-effective. Now that she had her analysis, it was time for Maria to share her recommendation so the company could make a data
driven decision. She summarized her results using clear and compelling
visuals of the analysis. This helped her stakeholders understand the solution
to the original problem. Finally, Anywhere Gaming
Repair took action, they worked with a local
podcast production agency to create a 30 second ad
about their services. The ad ran on podcast for
a month, and it worked. They saw an increase in customers after just
the first week. By the end of week 4, they had 85 new
customers. There you go. Effective problem solving using data analysis phases in action. Now, you’ve seen
how the six phases of data analysis can
be applied to problem solving and how you can use that to solve real world problems.

I’m Nikki and I
manage the education, evaluation, assessment,
and research team. My favorite part of the data
analysis process is finding the hardest problem and asking a million questions about it and seeing if it’s even
possible to get an answer. One of the problems that
we’ve tackled here at Google is our Noogler
onboarding program, which is how we
onboard new hires. One of the things that we’ve
done is ask the question, how do we know whether or not Nooglers are
onboarding faster through our new
onboarding program than our old onboarding program
where we used to lecture them. We worked really closely with
the content providers to understand just exactly what does it mean to onboard
someone faster? Once we asked all the questions, what we did is we prepared the data by understanding who was the population of the new
hires that we were examining. We prepared our data by going through and understanding
who our populations were, by understanding who
our sample set was, who our control group was, who our experiment group was, where were our data sources, and make sure that it
was in a set, in a format that was clean and digestible for us to write the
proper scripts for. So the next step for us was to process the data to
make sure that it was in a format that we could
actually analyze in SQL, making sure that was
in the right format, in the right columns, and in the right tables for us. To analyze the data, we wrote scripts in SQL and
in R to correlate the data to the control group or
the experiment group and interpret the
data to understand, were there any changes in the behavioral
indicators that we saw? Once we analyze all the data, we want to report on it in a way that our stakeholders
could understand. Depending on who our
stakeholders were, we prepared reports, dashboards and presentations, and shared
that information out. Once all of our
reports were complete, we saw really positive
results and decided to act on it by continuing our project-based learning
onboarding program. It was really satisfying to
know that we have the data to support it and that
it really, really worked. And not just that the data was there, but that we knew
that our students were learning and that
they were more productive, faster back on their jobs.

Data analysis is the process of inspecting, cleaning, transforming, and modeling data with the goal of discovering useful information, informing conclusions, and supporting decision-making.

There are many different types of data analysis problems, but some of the most common include:

• Missing data: This is a problem that occurs when some of the data is missing from a dataset. This can be caused by a variety of factors, such as human error, technical problems, or deliberate deletion. Missing data can make it difficult to analyze the data and draw accurate conclusions.
• Outliers: These are data points that are significantly different from the rest of the data in a dataset. Outliers can be caused by a variety of factors, such as measurement errors, data entry errors, or unusual events. Outliers can distort the results of data analysis, so it is important to identify and address them.
• Duplicate data: This is a problem that occurs when there are two or more data points that are exactly the same. Duplicate data can make it difficult to analyze the data and draw accurate conclusions.
• Inconsistent data: This is a problem that occurs when the data in a dataset is not consistent. This can be caused by a variety of factors, such as using different data collection methods, using different data formats, or entering data incorrectly. Inconsistent data can make it difficult to analyze the data and draw accurate conclusions.
• Corrupt data: This is a problem that occurs when the data in a dataset is damaged or corrupted. This can be caused by a variety of factors, such as hardware problems, software problems, or human error. Corrupt data can make it impossible to analyze the data and draw any conclusions.

These are just some of the most common problem types in data analysis. There are many other types of problems that can occur, and the specific problems that you encounter will depend on the specific dataset that you are working with.

To address these problems, data analysts use a variety of techniques, such as data cleaning, data imputation, and data normalization. Data cleaning is the process of identifying and correcting errors in the data. Data imputation is the process of filling in missing data. Data normalization is the process of transforming the data so that it is all on the same scale.

By addressing these problems, data analysts can ensure that the data is accurate and reliable, and that the results of data analysis are valid and useful.

Here are some additional tips for dealing with common problem types in data analysis:

• Be aware of the different types of problems that can occur.
• Use appropriate techniques to address the problems.
• Document your work so that you can track your progress and troubleshoot problems.
• Get help from a more experienced data analyst if you are stuck.

By following these tips, you can effectively deal with common problem types in data analysis and produce accurate and reliable results.

The finding patterns problem type could involve which of the following actions?

Using historical data to create a report that shows when batteries have been replaced on critical equipment

The finding patterns problem type could involve using historical data to create a report that shows when batteries on critical equipment have been replaced. Historical patterns can be used to help implement proper maintenance to prevent battery failures in the future.

In a previous video, I shared how data
analysis helped a company figure out where to
advertise its services. An important part
of this process was strong
problem-solving skills. As a data analyst, you’ll find that problems
are at the center of what you do every single day, but that’s a good thing. Think of problems as opportunities
to put your skills to work and find creative
and insightful solutions. Problems can be small or large, simple or complex, no problem is like another
and they all require a slightly different approach but the first step
is always the same: Understanding what kind of
problem you’re trying to solve and that’s what we’re
going to talk about now. Data analysts work with
a variety of problems. In this video, we’re going to
focus on six common types. These include: making
predictions, categorizing things, spotting something unusual,
identifying themes, discovering connections,
and finding patterns. Let’s define each of these now. First, making predictions. This problem type involves
using data to make an informed decision about how things may be in the future. For example, a hospital
system might use a remote patient monitoring to predict health events for
chronically ill patients. The patients would take their health vitals
at home every day, and that information combined
with data about their age, risk factors, and other
important details could enable the hospital’s
algorithm to predict future health problems and even reduce future
hospitalizations. The next problem type
is categorizing things. This means assigning
information to different groups or clusters
based on common features. An example of this
problem type is a manufacturer that
reviews data on shop floor employee performance. An analyst may create
a group for employees who are most and least
effective at engineering. A group for employees
who are most and least effective at repair
and maintenance, most and least
effective at assembly, and many more
groups or clusters. Next, we have spotting
something unusual. In this problem type, data analysts identify data that is different from the norm. An instance of
spotting something unusual in the real world is a school system that has a sudden increase in the
number of students registered, maybe as big as a 30 percent jump in
the number of students. A data analyst might look into this upswing and discover that several new apartment
complexes had been built in the school
district earlier that year. They could use this analysis
to make sure the school has enough resources to handle
the additional students. Identifying themes is
the next problem type. Identifying themes takes
categorization as a step further by grouping information
into broader concepts. Going back to our
manufacturer that has just reviewed data on the
shop floor employees. First, these people are
grouped by types and tasks. But now a data analyst could take those categories
and group them into the broader concept of low productivity and
high productivity. This would make it possible
for the business to see who is most and
least productive, in order to reward
top performers and provide additional support to those workers who
need more training. Now, the problem type of
discovering connections enables data analysts to find similar challenges faced
by different entities, and then combine data and
insights to address them. Here’s what I mean; say a scooter company
is experiencing an issue with the wheels it
gets from its wheel supplier. That company would have to
stop production until it could get safe, quality
wheels back in stock. But meanwhile, the wheel
companies encountering the problem with the rubber
it uses to make wheels, turns out its rubber
supplier could not find the right
materials either. If all of these entities
could talk about the problems they’re facing
and share data openly, they would find a lot of similar challenges
and better yet, be able to collaborate
to find a solution. The final problem type
is finding patterns. Data analysts use data to find patterns by using
historical data to understand what happened in the past and is therefore
likely to happen again. Ecommerce companies use data to find patterns all the time. Data analysts look at
transaction data to understand customer buying habits at certain points in time
throughout the year. They may find that
customers buy more canned goods right
before a hurricane, or they purchase fewer
cold-weather accessories like hats and gloves
during warmer months. The ecommerce companies can use these insights to make sure they stock the right amount of products at these key times. Alright, you’ve now learned six
basic problem types that data analysts
typically face. As a future data analyst, this is going to be valuable
knowledge for your career. Coming up, we’ll talk a bit more about these problem
types and I’ll provide even more
examples of them being solved by data analysts. Personally, I love
real-world examples. They really help me better
understand new concepts. I can’t wait to share even more actual cases
with you. See you there.

Data analysis is the process of inspecting, cleaning, transforming, and modeling data with the goal of discovering useful information, informing conclusions, and supporting decision-making.

In the real world, data analysts often face a variety of problems, including:

• Data quality: Data quality refers to the accuracy, completeness, and consistency of data. Data analysts need to ensure that the data they are working with is of high quality in order to produce accurate and reliable results.
• Data volume: The amount of data that is available is constantly increasing. This can make it difficult to manage and analyze large datasets.
• Data complexity: Data can be complex and heterogeneous. This means that it can be difficult to understand and analyze data that comes from different sources in different formats.
• Data privacy and security: Data analysts need to be aware of the privacy and security risks associated with data analysis. They need to take steps to protect the confidentiality, integrity, and availability of data.
• Interpretation of results: The results of data analysis can be complex and difficult to interpret. Data analysts need to be able to communicate the results of their analysis in a clear and concise way.

These are just some of the problems that data analysts face in the real world. By understanding these problems, data analysts can develop strategies to address them and produce accurate and reliable results.

Here are some additional tips for dealing with problems in data analysis in the real world:

• Use appropriate tools and techniques.
• Collaborate with other data analysts and stakeholders.
• Document your work.
• Get help from a more experienced data analyst if you are stuck.

By following these tips, you can effectively deal with problems in data analysis in the real world and produce accurate and reliable results.

Here are some specific examples of problems in data analysis in the real world:

• A retailer may want to analyze customer data to identify trends in purchasing behavior. However, the data may be incomplete or inaccurate, making it difficult to draw accurate conclusions.
• A healthcare provider may want to analyze patient data to identify potential risks for disease. However, the data may be complex and heterogeneous, making it difficult to understand and analyze.
• A financial institution may want to analyze financial data to identify fraud. However, the data may be sensitive and require special security measures.

These are just a few examples of the many problems that data analysts face in the real world. By understanding these problems, data analysts can develop strategies to address them and produce accurate and reliable results.

You’ve been learning about six common
problem types of data analysts encounter, making predictions, categorizing things,
spotting something unusual, identifying themes, discovering
connections, and finding patterns. Let’s think back to our real world
example from a previous video. In that example, anywhere gaming repair wanted to figure
out how to bring in new customers. So the problem was, how to determine
the best advertising method for anywhere gaming repair’s target audience. To help solve this problem,
the company used data to envision what would happen if it
advertised in different places. Now nobody can see the future but
the data helped them make an informed decision about how things
would likely work out. So, their problem type
was making predictions. Now let’s think about the second
problem type, categorizing things. Here’s an example of a problem
that involves categorization. Let’s say a business wants to improve
its customer satisfaction levels. Data analysts could review recorded
calls to the company’s customer service department and evaluate
the satisfaction levels of each caller. They could identify certain key words or
phrases that come up during the phone calls and then assign them
to categories such as politeness, satisfaction, dissatisfaction,
empathy, and more. Categorizing these key words gives
us data that lets the company identify top performing customer
service representatives, and those who might need more coaching. This leads to happier customers and
higher customer service scores. Okay, now let’s talk about a problem that
involves spotting something unusual. Some of you may have a smart watch,
my favorite app is for health tracking. These apps can help people stay healthy by
collecting data such as their heart rate, sleep patterns,
exercise routine, and much more. There are many stories out there
about health apps actually saving people’s lives. One is about a woman who was young,
athletic, and had no previous medical problems. One night she heard
a beep on her smartwatch, a notification said her
heart rate had spiked. Now in this example think of
the watch as a data analyst. The watch was collecting and
analyzing health data. So when her resting heart rate was
suddenly 120 beats per minute, the watch spotted something unusual
because according to its data, the rate was normally around 70. Thanks to the data her smart watch gave
her, the woman went to the hospital and discovered she had a condition which
could have led to life threatening complications if she hadn’t
gotten medical help. Now let’s move on to the next type
of problem: identifying themes. We see a lot of examples of this
in the user experience field. User experience designers study and work to improve the interactions people
have with products they use every day. Let’s say a user experience designer
wants to see what customers think about the coffee maker his
company manufactures. This business collects anonymous
survey data from users, which can be used to answer this question. But first to make sense of it all, he will need to find themes that
represent the most valuable data, especially information he can use to
make the user experience even better. So the problem the user experience
designer’s company faces, is how to improve the user experience for
its coffee makers. The process here is kind of
like finding categories for keywords and phrases in
customer service conversations. But identifying themes goes even
further by grouping each insight into a broader theme. Then the designer can pinpoint
the themes that are most common. In this case he learned users often
couldn’t tell if the coffee maker was on or off. He ended up optimizing the design with
improved placement and lighting for the on/off button, leading to the product
improvement and happier users. Now we come to the problem
of discovering connections. This example is from
the transportation industry and uses something called
third party logistics. Third party logistics partners
help businesses ship products when they don’t have their own trucks,
planes or ships. A common problem these partners face is
figuring out how to reduce wait time. Wait time happens when a truck driver
from the third party logistics provider arrives to pick up a shipment but
it’s not ready. So she has to wait. That costs both companies time and
money and it stops trucks from getting back on
the road to make more deliveries. So how can they solve this? Well, by sharing data the partner
companies can view each other’s timelines and see what’s causing
shipments to run late. Then they can figure out how to
avoid those problems in the future. So a problem for one business doesn’t
cause a negative impact for the other. For example, if shipments are running late
because one company only delivers Mondays, Wednesdays and Fridays, and the other
company only delivers Tuesdays and Thursdays, then the companies can choose
to deliver on the same day to reduce wait time for customers. All right, we’ve come to our final
problem type, finding patterns. Oil and gas companies are constantly
working to keep their machines running properly. So the problem is,
how to stop machines from breaking down. One way data analysts can do
this is by looking at patterns in the company’s historical data. For example, they could investigate
how and when a particular machine broke down in the past and then generate
insights into what led to the breakage. In this case, the company saw pattern
indicating that machines began breaking down at faster rates when maintenance
wasn’t kept up in 15 day cycles. They can then keep track
of current conditions and intervene if any of these
issues happen again. Pretty cool, right? I’m always amazed to hear about
how data helps real people and businesses make meaningful change. I hope you are too. See you soon.

Hi, I’m Anmol. I’m the Head of Large
Advertiser Marketing Analytics within the Marketing
Team at Google. At its core, my job
is about connecting the right user with the right
message at the right time. The first step is really to get a broad sense of the certain
pattern that’s occurring. So for example, we know that
this particular segment of users is more responsive
to this type of content. Once we’re able to actually see this hypothesis through the data, we do testing to ensure that the hypothesis
is actually correct. So for example, we would test sending these pieces of content
to this segment of users, and actually verify within
a controlled environment whether that response rate is actually higher for
that type of content, or whether it isn’t. Once we’re able to actually
verify that hypothesis, we go back to the stakeholders, in this case, our
marketers, and say, we’ve proven within a relatively
high degree of certainty that this particular
segment is more responsive to this type of
content, and because of that, we’re recommending that you produce more of this
type of content. Our stakeholders
really get to see the whole evolution from
hypothesis to proven concept, and they’re able to come with us on the journey
on how we’re proving out these hypotheses and then
eventually turning them into strategies and
recommendations for the business. The outcome in this
case was that we were able to actually change the way our whole marketing team worked to actually make it
much more user-centric. Instead of, from our perspective, coming up with content that
we think the users need, we’re actually going in
the other direction of figuring out what
users need first, proving that they need certain things or they
don’t need certain things, and then using that information
going back to marketers and coming up with content
that fulfills their need. So it really changed
the direction of how we produce things.

SMART questions are specific, measurable, achievable, relevant, and time-bound. They are used to guide data analysis and ensure that the analysis is focused and productive.

Here are the five characteristics of SMART questions:

• Specific: SMART questions are specific and unambiguous. They should be clear about what information is being sought.
• Measurable: SMART questions should be measurable. This means that the answer to the question can be quantified.
• Achievable: SMART questions should be achievable. The answer to the question should be possible to obtain with the available data and resources.
• Relevant: SMART questions should be relevant to the overall goal of the data analysis.
• Time-bound: SMART questions should have a deadline. This will help to ensure that the analysis is completed in a timely manner.

Here are some examples of SMART questions in data analysis:

• Specific: What is the average customer lifetime value for our product?
• Measurable: How many customers have churned in the past month?
• Achievable: Can we increase website traffic by 10% in the next quarter?
• Relevant: How does the average customer journey differ between new and returning customers?
• Time-bound: Can we reduce the time it takes to process a customer order by 50% in the next year?

By asking SMART questions, data analysts can ensure that their analysis is focused, productive, and relevant to the overall goal.

Here are some tips for asking SMART questions:

• Start by identifying the specific information you are seeking.
• Be clear about what data you have available and what resources you have.
• Consider the overall goal of the data analysis and make sure your questions are relevant.
• Set a deadline for answering your questions.

By following these tips, you can ask SMART questions that will help you get the most out of your data analysis.

Questions leading to answers that can be quantified and assessed align with which component of the SMART methodology?

Measurable

Questions leading to answers that can be quantified and assessed align with the measurable component of the SMART methodology.

While considering a research question, a data analyst follows the SMART methodology. They limit their analysis to include data from July 2012 to August 2012. What component of the SMART framework describes this decision?

Time-bound

Limiting analysis to a certain time period describes time-bound questions. They help limit the range of analysis possibilities and enable data analysts to focus on the most relevant data.

Now that we’ve talked about
six basic problem types, it’s time to start solving them. To do that, data analysts start
by asking the right questions. In this video, we’re going to learn
how to ask effective questions that lead to key insights you can use
to solve all kinds of problems. As a data analyst,
I ask questions constantly. It’s a huge part of the job. If someone requests that
I work on a project, I ask questions to make sure we’re on the
same page about the plan and the goals. And when I do get a result, I question it. Is the data showing me
something superficially? Is there a conflict somewhere
that needs to be resolved? The more questions you ask,
the more you’ll learn about your data and the more powerful your insights
will be at the end of the day. Some questions are more
effective than others. Let’s say you’re having
lunch with a friend and they say, “These are the best
sandwiches ever, aren’t they?” Well, that question doesn’t
really give you the opportunity to share your own opinion, especially if you happen to disagree and
didn’t enjoy the sandwich very much. This is called a leading question
because it’s leading you to answer in a certain way. Or maybe you’re working on a project and
you decide to interview a family member. Say you ask your uncle,
did you enjoy growing up in Malaysia? He may reply, “Yes.” But you haven’t learned much
about his experiences there. Your question was closed-ended. That means it can be
answered with a yes or no. These kinds of questions rarely
lead to valuable insights. Now what if someone asks you,
do you prefer chocolate or vanilla? Well, what are they
specifically talking about? Ice cream, pudding,
coffee flavoring or something else? What if you like chocolate ice cream but
vanilla in your coffee? What if you don’t like either flavor? That’s the problem with this question. It’s too vague and lacks context. Knowing the difference between effective
and ineffective questions is essential for your future career as a data analyst. After all, the data analyst
process starts with the ask phase. So it’s important that we
ask the right questions. Effective questions
follow the SMART methodology. That means they’re specific, measurable,
action-oriented, relevant and time-bound. Let’s break that down. Specific questions are simple,
significant and focused on a single topic or
a few closely related ideas. This helps us collect information that’s
relevant to what we’re investigating. If a question is too general, try to narrow it down by
focusing on just one element. For example, instead of asking
a closed-ended question, like, are kids getting enough
physical activities these days? Ask what percentage of kids
achieve the recommended 60 minutes of physical activity
at least five days a week? That question is much more specific and
can give you more useful information. Now, let’s talk about
measurable questions. Measurable questions can be quantified and
assessed. An example of an unmeasurable question
would be, why did a recent video go viral? Instead, you could ask how many times
was our video shared on social channels the first week it was posted? That question is measurable because
it lets us count the shares and arrive at a concrete number. Okay, now we’ve come to
action-oriented questions. Action-oriented questions
encourage change. You might remember that problem solving
is about seeing the current state and figuring out how to transform
it into the ideal future state. Well, action-oriented
questions help you get there. So rather than asking, how can we get customers to
recycle our product packaging? You could ask, what design features will
make our packaging easier to recycle? This brings you answers you can act on. All right,
let’s move on to relevant questions. Relevant questions matter, are important and have significance to
the problem you’re trying to solve. Let’s say you’re working on a problem
related to a threatened species of frog. And you asked, why does it matter that Pine Barrens
tree frogs started disappearing? This is an irrelevant
question because the answer won’t help us find a way to prevent
these frogs from going extinct. A more relevant question would be, what
environmental factors changed in Durham, North Carolina between 1983
and 2004 that could cause Pine Barrens tree frogs to disappear
from the Sandhills Regions? This question would give us answers
we can use to help solve our problem. That’s also a great example for
our final point, time-bound questions. Time-bound questions specify
the time to be studied. The time period we want
to study is 1983 to 2004. This limits the range of possibilities and enables the data analyst
to focus on relevant data. Okay, now that you have a general
understanding of SMART questions, there’s something else
that’s very important to keep in mind when crafting questions,
fairness. We’ve touched on fairness before,
but as a quick reminder, fairness means ensuring that your
questions don’t create or reinforce bias. To talk about this,
let’s go back to our sandwich example. There we had an unfair question because it
was phrased to lead you toward a certain answer. This made it difficult to answer honestly
if you disagreed about the sandwich quality. Another common example of an unfair
question is one that makes assumptions. For instance, let’s say a satisfaction survey is given
to people who visit a science museum. If the survey asks,
what do you love most about our exhibits? This assumes that the customer loves
the exhibits which may or may not be true. Fairness also means crafting questions
that make sense to everyone. It’s important for
questions to be clear and have a straightforward wording
that anyone can easily understand. Unfair questions also can make your
job as a data analyst more difficult. They lead to unreliable feedback and missed opportunities to gain
some truly valuable insights. You’ve learned a lot about how
to craft effective questions, like how to use the SMART framework
while creating your questions and how to ensure that your questions
are fair and objective. Moving forward,
you’ll explore different types of data and learn how each is used to
guide business decisions. You’ll also learn more about
visualizations and how metrics or measures can help create success. It’s going to be great!

[MUSIC]
Hi, I’m Evan. I’m a learning portfolio
manager here at Google, and I have one of the coolest
jobs in the world where I get to look at all the different
technologies that affect big data and then work them into training courses
like this one for students to take. I wish I had a course like this
when I was first coming out of college or high school. It was honestly a data analyst course
that’s geared in the way like this one is if you’ve already taken
some of the videos really prepares you to do
anything you want. It will open all of those doors that you want for any of those roles inside
of the data curriculum. Well, what are some of those roles? There are so many different career paths
for someone who’s interested in data. Generally, if you’re like me, you’ll come in through the door as a data
analyst maybe working with spreadsheets, maybe working with small,
medium, and large databases, but all you have to remember
is 3 different core roles. Now there’s many in special, whether
specialties, within each of these different careers, but
these three are the data analysts, which is generally someone who
works with SQL, spreadsheets, databases, might work as
a business intelligence team creating those dashboards. Now where does all that data come from? Generally, a data analyst will
work with a data engineer to turn that raw data into actionable pipelines. So you have data analysts,
data engineers, and then lastly, you might have data scientists who
basically say the data engineers have built these beautiful pipelines. Sometimes the analyst do that too.
The analysts have provided us with clean and actionable data. Then the data scientists then worked actually to turn
it into really cool machine learning models or
statistical inferences that are just well beyond anything you
could have ever imagined. We’ll share a lot of resources in links for
ways that you can get excited for each of these different roles. And the best part is, if you’re like me when I went into school, I didn’t
know what I wanted to do and you don’t have to know at the outset
which path you want to go down. Try ’em all. See what you really, really like. It’s very personal. Becoming
a data analyst is so exciting. Why?
Because it’s not just like a means to an end. It’s just taking a career path where so
many bright people have gone before and have made the tools and technologies
that much easier for you and me today. For example, when I was starting
to learn SQL or the structured query language that you’re going to
be learning as part of this course, I was doing it on my local laptop and
each of the queries would take like 20, 30 minutes to run and
it was very hard for me to keep track of different SQL
statements that I was writing or share them with somebody else. That
was about 10 or 15 years ago. Now, through all
the different companies and all the different tools that
are making data analysis tools and technologies easier for you, you’re going
to have a blast creating these insights with a lot less of the overhead that
I had when I first started out. So I’m really excited to
hear what you think and what your experience is going to be.