Statistics

Are you interested in maths? Good, then statistics is for you. The theoretical track of statistics is concerned with deriving optimal methods for statistical problems. To put it differently, someone must develop the mathematical machinery that computer software uses. This could involve integration theory, matrix algebra, differential equations etc.

Are you deterred by maths? Then statistics is for you. If you choose to work with applied statistical analysis (i.e. analyzing real data), a portion of common sense and some handiness with computer software is usually enough for statistical data analysis.

Do you enjoy working with others? Good, then statistics if for you. Skills in statistical data analysis is required by researchers in for example, economics, business administration, medicine, biology, engineering and many other fields. Those trained in statistical data analysis are therefore attractive project partners and frequently invited to their team.

More about a prospective career in statistics

Data scientist/analyst

Spotify

• Perform analysis on large sets of data to extract impactful insights on user behavior that will help drive product and design decisions
• Consumer experience: understand how users curate content

Ikea: Franchise, Range and Supply

• Supporting business stakeholders in understanding the potential of getting data to actionable information to enable business movement

Finance/sales analyst

Swedbank – Savings & Investments business

• Help the organisation to make factual based decisions based on data analyses and follow-up of sales, margins and customer/advisor activity

Analytical consultant

Google

• Large customer sales - working with customers, using online media with communication skills and analytical abilities. Partner with data engineers and data specialists to build processes, templates, and scripts

AI Specialist/Scientist

Meta

• Producing new science to understand intelligence and technology to make computers more intelligent

AstraZeneca

• Data, Analytics & AI group for Operations IT spanning various business areas such as Develop, Manufacture, Supply Chain, Quality and Procurement

Is statistics related to probability?

According to Encyclopedia Britannica, “probability and statistics (are) the (two) branches of mathematics concerned with the laws governing random events, including the collection, analysis, interpretation, and display of numerical data. Roughly, one may say that while statistics is concerned with questions such as “what probabilistic model might have generated this outcome (data)”, probability is concerned with the reversed question “what outcome (data) can we expect this probabilistic model to generate”.

Crystal clear? No?

Okay, let's try to make sense of this by using a simple example: Suppose we are going to toss a symmetric coin 100 times. What is the probability that we observe “heads” less than 30 times? Elementary probabilistic tools can help us calculate this number (the answer is, by the way, XX). So far, we are only involving probability theory, we haven’t actually involved any data in the calculation. Now let’s formulate another question: Suppose we toss a coin 100 times and get “heads” 30 times. Then we may conclude that, for this specific coin, the probability of observing “heads” is about 0.3. The fundamental difference lies in that the first cases start off with the assumption that the probability of heads=0.5, whereas the second case only relies on observations (data).

What are the differences between statistics, mathematical statistics and data science?

Statistics and mathematical statistics partly developed from different scientific branches and come to overlap each other significantly. It is sometimes said that mathematical statistics has traditionally had a stronger focus on probability theory and applications in the natural sciences while statistics has focused on statistical inference and applications in economics and social sciences. Naturally, the two subjects are closely related to data science.

Why all these Greek letters?

Data analysis involves lots of quantities of known or unknown value. When writing scientific papers or reports, it becomes impractical to assign explicit names to variables and parameters. By tradition, we use latin letters to denote variables (e.g., x=production costs, y=labour costs) while greek letters are reserved for parameters (e.g., µ=average annual rainfall, r=correleation between x and y).

Environmental research

• Statistics monitors the environment
• Investigates the natural environment and its interaction with humanity
• Monitors water and rivers, pollution, meteorology, and does earthquake risk analysis
• Describes characteristics, tests hypotheses, and makes predictions

Epidemiology

• Analyses how the pattern of disease causes change in the function of human beings
• Decision-making based on risk factors and preventive healthcare seeking
• Collecting and analysing data, to interpret results and draw conclusions
• Used in clinical research and public health

Finance

• Applies stochastic processes and nonlinear dynamics to financial markets
• Uses empirical studies to investigate financial time series
• Makes models for pricing derivatives and to predict stock movement
• Wants to understand behaviour of the changes in the stock market

Machine learning

• The study of computer algorithms that can improve automatically through experience and by the use of data
• Uses sample data to learn how to make predictions
• Uses computational statistics – transforming raw data into knowledge

Psychology

• Quantitative psychology – measures cognitive abilities by using mathematical modelling and statistical analysis
• Psychometrics – theory and technique of measurement of latent constructs (intelligence, disorders, achievements) that are not directly observed

Quality control

• Assures quality of all factors involved in production
• Inspection

Sports

• Statistical Science applied to sports