The ability to discern and utilize various data sources is crucial. Reliable data can come from government databases, reputable organizations, or through APIs that provide real-time data. Understanding the nuances of data formats is equally important; for instance, spreadsheets are excellent for tabular data, while databases are suited for larger, structured data sets. Mastery of these tools and the data they hold can provide invaluable insights, driving informed decisions and compelling narratives in your work. Personal experience with these tools can also add a unique perspective to your analysis, making your insights stand out.

Statistical literacy: Journalists need to understand basic statistical concepts such as averages, percentages, correlations, and significance levels to interpret data accurately and draw meaningful conclusions. Data sourcing and evaluation: Knowing where to find reliable datasets and how to assess their quality is crucial. This involves understanding data collection methods, potential biases, and limitations. Data cleaning and preprocessing: Often, raw data needs to be cleaned and prepared for analysis. Journalists should be proficient in data cleaning techniques to ensure accuracy and reliability in their reporting.

1. Data Source Identification: The ability to locate reliable and relevant data sources for journalistic stories. 2. Data Evaluation: The skill to assess the type, quality, and origin of data, ensuring its validity and credibility. 3. Data Format Familiarity: Knowledge of different data formats (spreadsheets, databases, APIs, web scraping tools) and how to access and manipulate them. 4. Data Analysis and Visualization: The capability to analyze data, extract insights, and present findings effectively through visualizations.

Data sources can include government databases, academic studies, surveys, and online datasets, while formats can vary from CSV files to JSON files, and from spreadsheets to relational databases. Additionally, data sources can also include social media platforms, sensor data, and IoT devices, while formats can include images, audio, and video files.

There are a few big "buckets" to look at when using data for reporting. 1. Source - you've got to trust the source of the data, and if you're getting conflicting numbers, you have to keep digging because that's a red flag. 2. Sample size for surveys - this is a really big one. Think of shampoo commercials that say "87% of 15 people said their hair was shinier". Small sample sizes and dramatic percentages are a red flag and you have to understand the importance of a sample size to get a statistically sound result. 3. Aggregation - The devil is in the detail. When you bundle different segments into one homogenous, convenient blob, the output you get might make ffor interesting reading, but it will lack nuance and context.

In my experience, it’s critical to balance technical proficiency and storytelling. While mastery of tools for data visualization is essential, they often fall short of conveying the deeper narrative behind the data. It's not just about presenting facts through graphs and charts; it's about weaving these elements into a coherent, engaging story. It’s what journalists do.

Equally important is understanding what your intent is in visualizing data. Know if you're adopting Exploratory or Explanatory positioning. Cole Nussbaumer Knaflic articulates this concept in her book Storytelling with Data - she delineates between these two concepts by saying "exploratory analysis is what you do to figure out whats noteworthy - explanatory analysis is when you have a specific thing you want to explain". Knowing what mode you're taking here will inform how you design your visualizations.

Specifically with tools like Tableau, I have found that it is easy to get carried away with what you can do in terms of visualisation, and forget that what you need to do in order to tell an effective story is actually much simpler. So "Keep it simple!" is a guiding principle for me. You can always make it as complex as it needs to be, but when speaking to a non-analyst, the simplicity of a data visual will lead to more engagement, more questioning and better understanding - which is what we usually aspire for.

Data without analysis is just noise. Just like music needs to be created by understanding the vibrations and the resulting sound waves, and then processing and organising these sound waves utilizing musical instruments to compose melodies and tell a captivating story through resulting music. Similarly, data needs to be organised, analysed and visualised in order to create a captivating data story that results in meaningful insights. The visualization of data is done by creating meaningful graphs, charts and tables that together tell the story you compose.

Data analysis involves using statistical techniques and data visualization tools to extract insights and patterns from data, while data visualization enables the presentation of complex data in a clear and concise manner. Common data analysis techniques include regression analysis, clustering, and hypothesis testing, while data visualization tools include Tableau, Power BI, and D3.js.

The ethics and security of data is paramount, especially with the growing concerns around privacy and the tightening privacy laws. Technology advancements have also brought out the ugly face of faking news, deepfake, and artificially manipulated data in form of doctored content - text, photos, or videos, using AI. The ethics and security also extends to protecting data against theft and copyright violations. Giving required credit for data borrowed from credible sources, adding watermark on content that has copyright, masking data that might violate someone's privacy are some ways of following proper data ethics and security practices.

I am reminded of a situation where an article was posted about a layoff, and the website not only posted the number of people laid off, but also the names of the individuals laid off! Really, really sloppy. At the very least, reasing up on GDPR (Europe) and CCPA (California) should help any data-journalist get a grounding on the right usage of information that can directly lead to a breach of privacy or could do so indirectly. When dealing with data on a server or a computer, there is onus on the person processing this data to ensure that it is locked up properly, and that only the right people have access to it.

Data ethics refers to the principles and practices that guide responsible data collection, use, and storage, while data security involves protecting data from unauthorized access, use, disclosure, disruption, modification, or destruction. Organizations must ensure that data is handled in a way that respects individuals' privacy and rights, while also implementing measures to prevent data breaches and cyber attacks.

It is critical to get beyond common, widely used solutions, particularly in processing unstructured data. Narrative Analytics should be first on the list. It's Journalism - so much of the 'data' is in the form of words & stories that flow in and around us. Most of the suggested tools above can not 'hear' or process narrative meaning. Or they slice them into pieces by employing sentiment scores, word vectors, or topic models which ignore context and have nothing to do with how people think. Narrative Analytics applies LLM's (not as a chatbot but as a listening array) to discover patterns, trends, and blindspots that other tools miss.

Proper functional knowledge of Data tools and platforms is important for effective data storytelling via analysis and visualizations. Using MS Word, PowerPoint and Excel is absolutely important for a journalist, and knowledge of BI tools like Google Looker studio, PowerBI, Tableau etc will be very useful in creating meaningful data stories and also for sharing and embedding onto the articles and other journalistic forums. Proper usage of CMS and their functionalities like Addons and reports is also important for a journalist in order to optimize the effectiveness of their articles. Apart from the above, functional knowledge of Search tools like Keyword research tools, Search Console, Lighthouse, and other SEO tools will also be very handy.

Data tools and platforms enable the collection, storage, processing, and analysis of data, with popular options including relational databases, NoSQL databases, and cloud-based data warehousing solutions. Data integration tools facilitate the consolidation of data from multiple sources, while data visualization tools help to present complex data insights in an intuitive and accessible manner. Big data platforms like Hadoop and Spark enable the processing and analysis of large-scale datasets, while machine learning platforms like TensorFlow and PyTorch support advanced analytics and predictive modeling.

Data literacy is not limited to data management and analysis. You need to be inquisitive and ask relevant questions. In order to understand what the data is saying, iterate upon production and content strategies based on actionable and insights gained from Data analysis. Engage with all stakeholders like Technology for on-page performance, Product on engagement related performance, Analytics team for all sorts of data insights, Monetization team to understand ROI on your efforts, and Business teams to understand overall business targets and goals. Its also important to know what is happening in the industry with competition analysis and participating in relevant forums and discussions.

The key mindset: be more ambitious about unstructured data. It's Journalism, so the (unstructured) views people express on an issue are essential data. Context is lost when data is boiled down to a multiple choice question, a sentiment score, or a keyword. People don't think in keywords. So it's essential to have the ambition to try emerging approaches in unstructured data, such as narrative analytics. Note, many analytics or quant researchers are less comfortable in this space, and disruptive technology like narrative analytics can create resistance among people whose expertise is based on legacy methods. Demand better, or find new experts.

A data-driven mindset and culture emphasize the importance of data-informed decision-making, encouraging organizations to base decisions on empirical evidence and analysis. This requires a shift in organizational culture, where data is seen as a valuable asset and stakeholders are empowered to use data to drive business outcomes. Data literacy is essential, with employees needing to understand how to collect, analyze, and interpret data to make informed decisions. By fostering a data-centric culture, organizations can drive innovation, improve operations, and achieve greater success.

Journalism is one domain where data misrepresentation is more dangerous than anything else. There are so many articles out where the data visualization is misleading e.g. the chart does not have a zero starting point which tweaks the truth. Representing & reading data is key when it comes to journalism.

In data journalism, it's important to frequently seek consumer feedback and constructively apply it to improve data visualization and analysis.

Correlation does not mean causation. The number of times I have seen people conflating the two is not funny. The root of most conspiracy theories is the lack of understanding of these two concepts. You have a huge responsibility as a journalist, please do justice to it.