Course homepage: here
For the purposes of this course Gender Data is defined as being "grounded in concepts and definitions that adequately reflect the diversity of women and men and capture all aspects of their lives."
Designed for gender advisors, data specialists, and other development professionals, Gender Data 101 Edition 2.0 is a 5-week blended online course featuring live events with gender and data experts that unpacks complex and practical considerations to understand, use, and share gender data with the ultimate goal of impact. Some of the exercises I complete(d) during the course will be available here for public viewing. The material will be organized by week and will reflect the activity I choose to complete (as oppose to all the rich available activies the course offers each week).
My hope is that by creating this repository 1) others can benefit from some of the engaging questions asked in the course and 2) I can record my own journey as an intentional and progressive data professional. Thank you for visiting and please feel free to reach out if you have any questions.
- Define best practices needed for gender data at all stages of the data life cycle: collection, processing, analysis, visualization, uptake, and impact.
- Identify the multiple forms of systemic discrimination that affect the overall efficacy of gender data
- Evaluate the limitations of the gender binary and how it may affect the phases of the gender data lifecycle
- Create action-oriented strategies and an intersectional approach to combat gender data inequities and biases
The framing of the activites and objectives will be quotes or paraphrased, so much of the framework here is the intellectual mentorship/property/paradigm of the course leaders. The answers to the activities are my work.
Description: The goal of this activity is to explore a dataset with sex-disaggregated data and answer a few questions.
Dataset Title: Electronic Police Report 2018
Source: Here
Description: All Police Reports filed by New Orleans Police Department officers including incident and supplemental reports containing the item number, location, disposition, signal, charges, offender race, offender gender, offender age, victim age, victim gender, and victim race. Police Reports can be updated when subsequent information is determined as a result of an investigation. In order to protect the privacy of victims, addresses are shown at the block level.
Date Last Updated: November 12, 2020
- What are the gender options?: (Example Answer 1: Gender options are binary, only M/F, Example Answer 2: Gender options are Male, Female, Other, etc.)
The gender options herein are Offender_Gender and Victim_Gender and are presented as a binary "MALE", "FEMALE", and blank fields without an option for "OTHER".
- Does the dataset have quantitative or qualitative data? Or both?
The dataset is primarily quantitative, but does contain some qualitative data (categorical data).
- Does the dataset take an intersectional approach? If so, to what extent?
The dataset does take an intersectional approach, but in a minimal way. The dataset contains some variables pertaining to RACE, ie. Offender_Race and Victim_Race, but these variables are not contextualized with any other quantitative (ex.average income in zipcode) or qualitative (ex. description of nuances in the dataset pertaining to these variables).
- Would you consider your dataset to be "gender data"? Why or why not?
I would not consider the dataset to be "gender data". "Gender Data" for the purposes of this course is defined as being "grounded in concepts and definitions that adequately reflect the diversity of women and men and capture all aspects of their lives." The course definition goes further to state that "collection methods for gender data take into account stereotypes, social and cultural factors that may induce bias." With all of that said, the dataset has the variables denoting sex, but lacks the social and cultural factors that may allow bias in. Nowhere in the dataset is there any information about class or the rate of "selective enforcement" officers choose to exercise, for example. Without any additional contextual information to nuance the dataset, I don't think the dataset crosses the threshold for "gender data".
- Considering your answers to questions 2 and 3, what does the dataset do well and what is it missing?
The dataset contains gender variables which is a positive in the dataset and also has disaggregated data which is even better for analysis. The dataset is missing additional social or cultural (read: "intersectional") data, be that qualitative or quantitative.
- What additional questions or comments you have for the dataset? For example: What methodologies were used to collect this data?
I would have liked to see more information concerning the income/salary/wage/employment data for the area in which the arrest took place or for the people involved in the incident to get a better sense of the economic context the issue took place in. The dataset could also be nuanced by adding factors like number/rate of homelessness/vagrancy in a zipcode, sexuality of those involved in an incident, number of children of those involved, and/or marriage status.
Description: Review the datasets (2) and take note of any areas of the data that may suffer from biases using terms from the "Examining Bias lesson section. Address the prompts.
- Dataset #1: Here
- Description: A local clinic's database system shut down. A clinic employee decided to gather data via Google Sheets instead of the usual secure system. This dataset was shared via email within clinic.
- Hypothetical Data Collector: Dr.Lopez
- Dataset #2: Here
- Description: A COVID-19 relief survey was sent via online link to households in a community in Chile. The link was sent during the weekend starting Friday at 5pm, to be completed by Monday. 142 households responded out 843 households surveyed.
- Hypothetical Data Collector: Rumy
- Summarize in a list the types of biases found in the datasets.
| Dataset | Type of Bias | Explained |
|---|---|---|
| #1 | Selection Bias | With respect to the "Race" category, "White" respondents are over represented in the dataset. Without any context on what community is being surveyed, I also think it is fair to say that "Black", "Asian", and "Other" racial groups are under represented. The survey results also do not appear to allow people to respond with multiple races. Given the history of settler colonialism, slavery, and war throughout much of the world, mutli-racial individuals/communities exist and should be represented in data. |
| #1 | Question Wording/Ordering Bias | With respect to the "Sex" category, there is no option for "Other" or "N/A" to represent lgbtqia+ respondents which may be using clinic services differently than "M" or "F" folks using clinic resources. |
| #2 | Social Desirability Response Bias | The question "Are you a legal resident?" has the ability to strike fear into the hearts of respondents and, more importantly, COVID-19 relief - the purpose of the survey - impacts citizens and non-citizens alike. This question may push respondents to respond "Yes" out of fear that "No" could make them targets while all they wanted to do was look for help in the midst of a pandemic. |
| #2 | Non-response bias | The question "Are you a legal resident?" has the ability to strike fear into the hearts of respondents and subsequently make them less likely to respond to the survey or less likely to answer certain questions on the survey (ex. some respondents did not respond to questions about their vaccination status). The question is the first one listed so it comes off as very important and is followed by address. This information in the wrong hands could result in unintended arrests, deportations, family seperation, and/or imprisonment. |
- Of the biases you found, which one(s) do you think are the most important to address first? Why?
Of the biases I found, the most important one to address is the selection bias issue from dataset #1. The bias in the dataset for "White" respondents could lead to erroneous strategy setting that does not represent the breadth of who uses, needs, or uses services up to certain price points. This bias is only made worse by the fact that sexual orientation is all together left out of the list of survey questions.
My concern is that the pattern of services used might be different between racial groups and/or subsects of different racial groups (ex. black men compared to black women, or hispanic lgbtqia+ compared to the broader lgbtqia+ community). By addressing this bias in the dataset, the local clinic in the hypothetical scenario might be able to better assist their community, find that their resources could be directed more towards the type of people actually coming, or that respondents tend to stop seeking care after a certain point.
- What is one method of consideration that Dr. Lopez and/or Rumy can apply during data collection to prevent the biases you found?
| Dataset | Type of Bias | How to Prevent |
|---|---|---|
| #1 | Selection Bias | The demographics of the communities served by the clinic and the demographics of the community at large need to be compared to the population that responded. Subsequently, if doing additional surveys are out of the question, one or more groups can be weighted so that they are represented more accurately for who uses services or could use their services. If additional surveys are possible, more surveys could be done with a new strategy on how to get underrepresented racial groups in the survey (ex. survey in additional languages or conducted live at strategic locations). |
| #1 | Question Wording/Ordering Bias | Replace with "Gender" and provide option for each of the following or an umbrella category for "Transgender", "Nonbinary", and/or "Prefer not to answer". |
| #2 | Social Desirability Response Bias | Remove the question about citizenship as it is not essential to the study or move it to later on in the survey after more important questions, COVID-19 related ones, have been answered to avoid bias. |
| #2 | Non-response Bias | Remove the question about citizenship as it is not essential to the study or move it to later on in the survey after more important questions, COVID-19 related ones, have been answered to avoid bias. |
- Are either of these datasets and example of gender data? Why or why not?
Unfortunately, I do not think either dataset #1 or #2 qualifies as gender data. For dataset #1, while it contains sex disaggregated data it does not contain gender disaggregated data. For dataset #2, the dataset does not collect sex or gender data let alone get into the nuances/context of the lives of people of different genders.
On a broader note, I think I am beginning to wonder what Gender Data might actually look like, aside from having a gender variable in a dataset. I'd like to see an example.
Description: Review and assess data from hypothetical COVID-19 related research projects(s). Subsequently, answer questions about the data, the process, and what way their analysis could be made better by adding a gender data analysis framework to it.
Dataset: The Sex, Gender and COVID-19 Project
Description: The COVID-19 Sex-Disaggregated Data Tracker is the world’s largest database of sex-disaggregated data on COVID-19. Dr. Lopez is primarily concerned with her own city's (Washington, D.C.) COVID-19 data. She has aggregated data from local nursing homes, hospitals, and clinics to analyze. We will help answer some of the questions the good doctor has about the data.
- Describe the metadata of this dataset (Where is the data sourced? If there are multiple sources, list just a few)
According to the course "Metadata includes general information about the dataset(s). It often includes definitions, rationale, methodology descriptions, and information about outliers." The Sex, Gender, and Covid-19 project has:
- status of sex disaggregated data broken down by country,
- status of covid-19 reporting by country,
- global level statistics on Covid-19 cases disaggregated by sex,
- and analyses of how sex and gender may or may not be included in policy directives around the world.
In summary, the project is an incredibly robust endeavor to collect, synthesize, and analyze sex disaggregated data from all around the globe. They also go on to use all of the data they have collected to push for more contries to collect and publicly report sex disaggregated data.
The sources used to produce this project include the following:
- Global Health 50/50 - is an independent, evidence-driven initiative to advance action and accountability for gender equality in global health.
- the African Population and Health Research Center - Is the continent’s premier progressive research institution and think tank, generating evidence to drive policy action to improve the health and wellbeing of African people. They focus on research, research capacity strengthening, and polcity engagement and communications.
- and the International Center for Research on Women - ICRW is a global research institute whose research evidence identifies women’s contributions as well as the obstacles that prevent them from being economically strong and able to fully participate in society. ICRW translates these insights into a path of action that honors women’s human rights, ensures gender equality and creates the conditions in which all women can thrive.
The project goes a step further by enabling its audience to review and download the data at the global level, country, and variable level.
- What specific data category should Dr. Lopez focus on and why? (For example: Men in Thailand and their COVID-19 death rate)
Dr. Lopez should focus on Washington, DC level/local level Covid-19 reporting that covers the lifecycle of covid from infection to mortality. The project only appears to have country level statistics and since her work is locally focused she may need a new data source all together.
- In what ways could a Gender Data Analysis Framework (GAF) help Dr. Lopez's analysis?
A GAF with the current data available in the project might not be as helpful to Dr.Lopez since it does not dig enough into her region of interest. However, even at this stage, a GAF may identify impediments to resources being distributed (ie. medical information and care) to women globally that she can later connect to the situation of women in Washington, D.C. when she does data particular to her area of interest.
In particular I think the Jhpiego's Gender Analysis Framework for Health Systems may be really useful. In the U.S. different communities have different beliefs about healthcare at large and COVID-19 in particular, so there may be a really meaningful correlation between what women in a community believe and how the rest of the community is performing COVID wise, acting as something of a bellweather for how the community may or may not respond to different vaccination drives or, dare I say, anti-vaccination drives.
- Watch the following video: Dr. Joel Selanikio on Truth, Lies, and Data Visualization in the Era of Coronavirus
- Read: An alternative to pink & blue: Colors for gender data
- Choose a visualization: UNHCR's COVID-19 deepens threats for displaced women and children
Answer the following Questions:
- What does the data visualization(s) do well? Many of the examples have more than one visualization, please be specific in your answer.
The visualizations do an incredible job of providing insight on the multiple layers/ranking of issues, and their respective scale, displaced women and children face; the "Most common child protection risks reported by UNHCR operations, 2016-2020" visualization is exemplary of this. This visualization enumerates a list of issues from seperation to child labour, violence, lack of documentation, etc. This layering/ranking puts the size of the problems clearly into view for the audience.
- Choose a framework: Tufte's Framework, Schwabish's Core Principles, Junk Chart's Trifecta, write 3-6 sentences about how the visualization incorporates or doesn't incorporate the elements of the framework.
Though not included in the list, I would like to talk about how the visualizations in the UNHCR's report do not address one of the core components/questions of the Harvard Analytical Framework, ie. who has access to what? I think the data visualizations could have somehow included numbers on how many countries restrict access to documentation, mobility (read: access to movement without supervision and vehicle licenses) , and/or avenues for legal recourse (read: ability to sue for rape, sexual assault, or abuse) for women. This may have fit in well with or closely following the statistics on "Most frequently reported child protection risks in 2020".
In the WorldBank Blogs, Gender Legal Expert Nayda L. Almodóvar-Reteguis writes about how some countries keep women from accessing and administering their own resources here and the Washington Post has talked about historical restrictions to women's movement [https://www.washingtonpost.com/news/worldviews/wp/2013/10/27/7-ridiculous-restrictions-on-womens-rights-around-the-world/]. These issues are well documented and widely known, so I think it is fair to want to include a layer to the analysis that addresses these impediments to helping these displaced women and children.
- What is one thing you could do to improve the visualization?
The "Demographics of focibly displaced people in 2020" visualization in the UN Refugee Agency's report on violence agaisnt displaced women and children has inconsistent and illogical age groupings which creates one particularly oversized group. Particularly, the age bucket for 18-59 should be broken up into young adults 18-29, middle aged 30-45, and older folks 46-59. These different age groupings, or something akin to them, would allow for a more granular picture of the forcible displaced communities as young adults and older folks may have different immediate and long term needs and solutions for their displacement, for example. Bucketing 18 through 59 year olds brushes over useful nuances.
Final Project Brainstorm:
At the moment, I think options 8. Reflection on Gender Data 101 course and 9. Create a Gender Data Visualization are my most likely options. Option 9 is the most exciting option to me, but due to time constraints, I think option 8 might be my best bet. Ultimately, I think I would really like to number 9 though and may hold out to do it.
My idea for a gender data visualization is to aggregate US GDP data or federal job number reports, and stats on domestic abuse and/or sexual violence over time. My hypothesis is that they are correlated and that down turns in the economy lead to upticks in intra-partner violence. The significance of this is in identifying thresholds for when we think changes in the economy may trigger violence at people's kitchen tables and so that we can subsequently get resources allocated accordingly. There may be better bell weather indicators for intra-partner violence, but as an economics major in uni I think the two are tied or can be tied in a research paper.
I will be making a project description for myself.
PRT 2: If applicable, review Dr. Lopez or Rumy's journey throughout our course. (review the 'Case Study' section in each week's modules)
Not applicable given that I will be creating a project descriptin for myself.
- Project Title:
"Cross Checking for Covid-19 Aid Eligibility, Federally and Statewide"
- Project Overview:
This project endeavors to lessen the administrative burdens of those looking for covid-19 relief.
Throughout the covid-19 pandemic, states and the federal government in the United States have rolled out various programs meant to help create a safety net for those experiencing any number of hardships. Some of the hardships have been caused by changes in employment (layoffs or decrease in hours), domestic violence, child food scarcity, mental health, and student learning loss. The patchwork of programs erected help address those hardships are administered by a variety of entities with different initial eligibility standards and subsequent requirements for continued assistance. This has created a somewhat kafkaesque process that those experiencing any number of the aforementioned hardships may find difficult to identify, apply to, balance, and track. This project intends to address the difficulties individuals face while going through the initial application and subsequent requirement processes.
- Goals and Objectives:
This project's top priority is to simplify the process of finding, filing for, and managing covid-19 assistance applications. The first goal is consolidating initial eligibility requirements and subsequent requirements across US state and federal programs. The second priority is leveraging submissions for one program to raise an alert for applicants to look into other programs that address similar or adjacent needs. In this priority, I am also thinking about the possibility of a user experience path with the option to submit eligibility documents to multiple programs of their choosing at the same time or submitting and having our platform identify programs whose criteria are met and following up to ask for any other necessary documents to. The third priority is to create a platform capable of distributing individual submissions meant to meet requirements to multiple programs on the behalf of applicants. The third and final priority is managing the stats around aid disbursed including amount, household type, and important timelines.
- Project Timeline:
I think this project will take a one to one and a half years to roll out for use by the general public.
- Methodology and Tools:
While I am unsure of how long the phases will take or which ones can be done simultaneously given that different , I think the following will need to be accomplished to roll this project out:
- research on the state and federal programs,
- cultivating partnerships with different offices to identify possible efficiencies,
- creating a user journey for one person
- architecting the backend to receive and distribute submitted documents and track relevant timelines and figures,
- creating a user experience contract that allows our program to act as an intermediary between applicants and the program,
- creating a partnership contract that ensures entities administering programs will accept our intermediary services on an extended basis,
- automating the submission of eligibility documents and notifying the applicant of upcoming verification steps or issues,
- architecting an intuitive user interface on the front end in different lagnguages and on different platforms,
- and, testing the project for durability by having an increasing number of fake users throughout the US on different platforms use different features at the same time.
Some of the tools I think will be needed to accomplish the priorities of this project are as follows:
- a website to facilitate data collection, application completion, and distribution,
- SQL / stored procedures to automate eligibility verification across programs once one initial program has been applied for, an operational warehouse capable of storing and cross verifying the multiple submissions of the potential 330 milion users, links between our tool and offices running programs to allow submissions made to our team to be sent to or shared live time with each program as seamlessly as possible, automated mobile notifications, and a calendar for each individual user that is automatically updated based on the programs they apply to.
- Which stakeholders should you, Dr. Lopez, or Rumy target?
Our stakeholders will be limited to people who are eligible for any of the programs we are able to create partnerships with.
- What do you think the project's value statement should be and why?
"Simplifying government aid applications - because the pandemic has made everything harder!"
This project's value statement should be the above because the priorities culminate in an easier experience for applicants.
- What is one expected project outcome?
One project outcome is helping applicants apply to multiple programs at the same time.
- What is the project feedback plan?
I think this project would take a two pronged approach to feedback. The first would be a standing space on the platform to take feedback. The second would be an automated helper on the right side of the screen where one option could be submitting impromptu feedback. All feedback should receive a "received message" at the least. Questions should be responded to or redirected to a FAQ page after some frequently asked questions have been established.
Brief Project Description: I intend to complete the reflection final project, project #8. While, in general, I learned a lot, to be more specific, I really appreciated the gender analysis frameworks and data viz principles this course covered.
At the beginning of the Gender Data 101 Edition 2.0 course, I was expecting to learn how to apply a gender based framework in the analysis and vizualization stages of the data lifecycle. However, at the outset of the course, the instructors made it very apparent that they intended demonstrate how a gender framework could be applied to each step of the lifecycle, from collection and processing, to analysis, to visualization, all the way through uptake and impact. I think this approach made the course more thorough and robust than initially anticipated - and I really enjoyed it all!
Of the various part of the program, I first and foremost enjoyed the material in the analysis and vizualization stages of the data lifecycle the most because they covered the stages of the process that I as a data analyst have the most exposure to and experience with. Second to that, I really enjoyed the videos and linked content to videos and external webpages that served as an example of the themes we were covering at a given time. And, finally, I really enjoyed rewatching the speakers who came to speak/present. Of the speakers, I think my favorites were Jonathan Schwabish of PolicyViz.com, and the speaker who addressed data privacy and collection work at a sexual assault survivor services organization.
I can confidently say that there was nothing missing from the course, and I wouldn't restructure the course. I can say this because the course's goals, estimated time requirements, and difficulty were laid out upfront. For a beginner course in gender data, this course was a well rounded self paced course with a number of avenues to explore based on one's learning style. I am really happy that a peer recommended that I take this course.
Terms and videos listed herein are from the TechChange Gender Data 101 course and are reproduced here to add to this record of what I learned during the course.
- Collection is the process of gathering and measuring information in a systematic fashion.
- Biases are personal and/or systematic judgments and pre-conceptions.
- Selection Bias, also known as sampling bias, is choosing a sample that is not representative of the population being studied. Selection bias also results when participants not being randomly sampled to a condition.
- Response Bias is a general term for any external factor that can influence a respondent’s answers.
- Non-response bias means certain members of the population being studied do not participate in the data capture. There are many reasons for non-responses bias. Examples include the medium of study (e.g. online survey) or time study is conducted (e.g. morning times).
- Demand characteristics are the ways participants of a study can influence the outcome.
- Social Desirability Response Bias is a type of response bias is when a respondent is influenced to provide answers that are socially desirable and are not necessarily their true responses.
- Question Wording/Ordering Bias occurs if the question is worded or ordered in such a way that it favors one response over another.
- Outlier/Extreme Response Bias occurs when there are extreme data values.
- Processing is the series of operations needed to prepare data for analysis.
- Raw data (also called dirty or unclean data) is data directly from the source.
- Incomplete data is data that lacks the key information.
- Duplicate data is data that repeats itself.
- Inaccurate data is information that is not correct.
- Inconsistent data is unstandardized data.
- Clean data is data that is accurate and formatted to a set of rules. These rules can be determined by your organization or by your data analysis needs. Data cleaning helps edit or remove incomplete, duplicate, incorrect, and inconsistent data.
- Data aggregation is systematically combining data from multiple sources together. Processing may or may not involve aggregation.
- Data security is the commitment to confidentiality and integrity of data.
- Personally Identifiable Information also known as PII is any data that can be used to distinguish or trace an individual's identity.
- Data encryption is the process of converting data into code to prevent unauthorized access.
- Intersectional Pay Equity by WeAllCount - Intersectional analysis demonstrated with Pay Equity example
- Metadata is information about the data. This information includes general information about the dataset(s). It often includes definitions, rationale, methodology descriptions, and information about outliers. Metadata usually appears in footnotes or near the dataset description.
- Ratio compares the frequency of one value for a variable with another value of the same variable.
- Rate a measurement of one value for a variable in relation to another measured quantity.
- Proportion number of times a particular value for a variable has been observed, by the total number of values in the population.
- Percentage expresses a value for a variable in relation to a whole population as a fraction of one hundred.
- Percentage Points used to express increments, drops, or differences. It represents decimal points. These are different from percentages.
- Mean the arithmetic average of a set of numbers or distribution. It is the most commonly used central tendency of a set of numbers. It is used for normal distributions and it is not robust, as it is influenced by outliers.
- Median the numeric value separating the higher half of a sample, a population, or a probability distribution, from the lower half. It is used for skewed distributions. It is computed by listing all numbers in ascending order and locating the number in the center of the distribution.
- Gender Analysis Frameworks or (GAFs) are a collection of tools for gender analysis.
- Jhpiego's Gender Analysis Framework for Health Systems - Jhiepgo's GAF categorizes the lives of women and men, girls and boys, into four main domains. All domains intersect with power. The four domains: access to assets, beliefs and perceptions, practices and participation, institutions, laws, and policies
- Oxfam's Feminist monitoring, evaluation, accountability and learning (MEAL) - Feminist Monitoring, Evaluation, Accountability and Learning (MEAL) is based on the understanding that transformative change in unequal gender and power relations is complex and non-linear. It challenges us to think differently about what is considered evidence, pushes the boundaries of how evidence is captured, questions who gives knowledge meaning and power, and promotes social transformation.
- Gender at Work Framework - Gender at Work Framework can be used by change agents to uncover opportunities and barriers to gender equality, map strategy for change, and guide evaluations of progress. The top two quadrants are related to the individual. The right quadrants are changes in individual conditions and the left is individual consciousness and capability. The bottom two quadrants are related to the system.
- Harvard Analytical Framework - The Harvard Analytical Framework, also known as the Gender Roles Framework, aims to identify the type and amount of work that men and women do in a household or community. The framework has three parts of inquiry: 1) Activity: Who does what? When? Where? 2) Access and Control: Who has access to what? Who has control over what? 3) Influencing Factors: What influences the division of labor and the access and control?
- Data visualization the graphic representation of data.
- Gender data visualization the graphic representation of gender data.
- Data Viz Framework ask yourself the following questions as you visualize data:
- Who is your viewer?
- What types of decisions does your viewer make? What information do they need? How will your data visualization add value?
- What is your viewer’s digital literacy?
- What is your viewer’s prior experience with data visualization?
- Tufte's Rules for data visualization, consider " analyz[ing] the datasets you have, first. Discover the story that it tells before you choose a visualization platform, which runs the risk of pre-specifying certain narratives of the data." Here are the ten rules Tufte outlines:
- Show Your Data
- Use Graphics
- Avoid Chartjunk
- Utilize Data-ink
- Use Labels
- Utilize Micro/Macro
- Separate Layers
- Use Multiples
- Utilize Color
- Understand Narrative
- Schwabish's Core Principles Jonathan Schwabish is a senior fellow at the Income and Benefits Policy Center at the Urban Institute. As an economist, he also specializes in data visualization and presentation design as a member of the communications team. He also hosts the PolicyViz Podcast, which focuses on data, open data, and data visualization.
- Show the data - The data is the most important part of the graph and should be presented in the clearest way possible. But that does not mean that all of the data must be shown—indeed, many graphs show too much.
- Reduce the clutter - Chart clutter, those unnecessary or distracting visual elements, tend to reduce effectiveness.
- Integrate the text and the graph - Visualizations should be constructed to complement the text and, at the same time, contain enough information to stand alone.
- Preattentive Processing - Effective data visualization taps into the brain’s preattentive visual processing.
- Junk Charts Trifecta The Junk Charts Trifecta Checkup is a general framework for data visualization criticism. It serves to organize the thinking behind data viz pieces by addressing three points of investigation:
- What is the QUESTION?
- What does the DATA say?
- What does the VISUAL say?
- Data Viz/ Visualization Tools
- Excel - Excel makes it super easy to store, analyze, clean, and quickly visualize data.
- Piktochart - A platform that generates beautiful, intuitive, and clean visuals for your data.
- Infogram - A web-based data visualization and infographics platform.
- Canva - A simplified graphic design tool and website that allows those of us without much design sense to pass as graphic artists.
- kepler.gl - high-performance web-based application for visual exploration of large-scale geolocation data sets. Built on top of deck.gl, kepler.gl can render millions of points representing thousands of trips and perform spatial aggregations on the fly.
- CARTO - cloud computing platform that provides GIS, web mapping, and spatial data science tools. The company is positioned as a location intelligence platform due to tools with an aptitude for data analysis and visualization that do not require previous GIS or development experience.
- What are Stakeholder Maps and why use them?
- How to create a Stakeholder Map
- Step 1 - Define the focus of the project
- Step 2 - Create a list of relevant stakeholders
- Step 3 - Prioritize the stakeholders
- Step 4 - Illustrate the stakeholders on the project map
- Step 5 - Sketch the value exchange between the stakeholders (ex. products, information, finances, etc.)
- Step 6 - Analyse your stakeholder map from different perspectives
- Step 7 - Save and document your Map