Data analysis thinking analysis methods and business knowledge -- analysis methods (II)

Analysis method

Comparative analysis

What is it? ？

• Compare several objects

  • The price of clothes
  • The big cup of milk tea 、 Middle cup 、 Small cup

What's the usage? ？

• Track whether there is a problem with the business

• A/B test

  • Make two versions of the same goal , Only one aspect is different , The others are the same , Then see the final effect , Finally, launch a better version

How to use ？

• With whom

  • And their
  • And Industry

• How to compare

  • Overall size of data

    • Average 、 Median 、 A business indicator

  • The volatility of the data as a whole

    • The value obtained by dividing the standard deviation by the average value is called the coefficient of variation （ Measure the volatility of the overall data ）

  • Trend change

    • From the perspective of time dimension, data changes over time

      • Time line chart

        • Take time as the horizontal axis 、 The data is the vertical axis

          • You can understand what changes have taken place in the data from the past to the present , You can also predict future trends through past changes

      • Chain ratio

        • Compared with the previous period , That is, different months of the same year

          • Data sets used to observe short-term

      • Year on year

        • Compared with the same period last year

          • Used to observe long-term data sets

matters needing attention

• The scale of the comparison object should be consistent

Hypothesis testing analysis method

What is it? ？

• logical reasoning

  • suggest a hypothesis
  • collect evidence
  • Come to the conclusion

What's the usage? ？

• Improve the ability of logical thinking
• Analyze the cause of the problem （ Attribution analysis ）

How to use ？

• From user 、 product 、 Competing products 3 Put forward assumptions in three dimensions

  • User problem

    • There is a problem with the channel

  • Product problems

    • Whether the server crashed
    • New product launch
    • A large number of complaints

  • Competitive products

    • Competitor promotions

• from 4P Marketing theory puts forward assumptions

  • product
  • Price
  • channel
  • Sales promotion

• Make assumptions from the business process

  • 1、 New users download APP, Then register

    • hypothesis 1： Low age users have no time to go to school during the day APP, Only play mobile phones on weekends

  • 2、 Users see the recommended content on the home page

    • hypothesis 2： Recommended content is not popular with younger users

  • 3、 stay APP Find what you are interested in

    • hypothesis 3： Younger users can't find their favorite content

matters needing attention

• The conclusion is not subjective conjecture , But rely on the evidence found to prove
• Hypothesis testing 3 Step is a process that needs to be repeated
• In the analysis method of using hypothesis verification , Other analytical methods are also used , For example, multi-dimensional disassembly analysis method
• In order to clarify the thinking , You can make a hypothesis test analysis chart , Put the question 、 hypothesis 、 The data is connected from top to bottom

Correlation analysis method

What is it? ？

• The relationship between various data

What's the usage? ？

• When studying the relationship between various data, or when something is affected by other factors
• Expand your thinking , Expand the field of vision from one data to multiple data
• Easy to understand

How to use ？

• Scatter plot

  • It can intuitively display the relationship between the two kinds of data

• The correlation coefficient

  • The size of the value indicates the degree of correlation between the two data

    • The greater the absolute value , The more relevant

      • 0~0.3 low correlation
      • 0.3~0.6 Moderate Correlation
      • 0.6~1 Highly correlated

  • Positive and negative reflect the direction of correlation between the two data

matters needing attention

• The difference between causality and Correlation

  • Causal relationship

    • A The occurrence of will lead to B happen ,B Because A

  • Correlation

    • A change ,B And change