{"id":59,"date":"2026-11-01T22:14:04","date_gmt":"2026-11-01T22:14:04","guid":{"rendered":"https:\/\/classprojects.digitalwcu.org\/HMAI\/methods-2\/"},"modified":"2026-11-09T06:22:48","modified_gmt":"2026-11-09T06:22:48","slug":"methods-2","status":"publish","type":"page","link":"https:\/\/classprojects.digitalwcu.org\/HMAI\/methods-2\/","title":{"rendered":"Methods"},"content":{"rendered":"\n<p class=\"has-foreground-color has-text-color has-link-color has-headings-font-family has-medium-large-font-size wp-elements-1 wp-block-paragraph\">Overview of the Methods (Assessed through documentation):<\/p>\n\n\n\n<ul class=\"wp-block-list has-headings-font-family\">\n<li><span style=\"text-decoration: underline;\">Statistics<\/span>: The researchers mention that they did not choose the sample size in advance, and that they did not randomize anything. It analyzes existing data and there was no experiment on participants.&nbsp;&nbsp;<\/li>\n\n\n\n<li><span style=\"text-decoration: underline;\">Data Set description<\/span>: The data was retrieved from IOS users of the Argus app, collected between July 2013 to December 2014. The obtained data like stepcount, location, demographics, and the data came from those who already had at least 10 days of data.&nbsp;<\/li>\n\n\n\n<li>V<span style=\"text-decoration: underline;\">erifying established physical activity trends:<\/span> In order to ensure that the data retrieved from the app was accurate, they initially whether it reproduces known patterns. For instance, they checked that activity falls with age and BMI, and checked against WHO data for correlations between activity, gender and obesity.&nbsp;<\/li>\n\n\n\n<li><span style=\"text-decoration: underline;\">Daily recorded steps and wear time: <\/span>since the steps are only counted when their phone is carried with them, the time between the first and last step of the day is used as a proxy for wear time. The average was 14 hours, however there is no correlation between longer wear time and more steps. Men were found to have longer wear times than women.&nbsp;<\/li>\n\n\n\n<li><span style=\"text-decoration: underline;\">Defining activity inequality:<\/span> measured with gini coefficient which is the same measure economists use to measure income inequality.&nbsp;<\/li>\n\n\n\n<li><span style=\"text-decoration: underline;\">Correlation between activity inequality and obesity: <\/span>correlated each country\u2019s activity inequality with obesity rate using r values and smoothed fitting curve called LOESS.&nbsp;<\/li>\n\n\n\n<li><span style=\"text-decoration: underline;\">Robustness of correlation between activity inequality and obesity:<\/span>\n<ul class=\"wp-block-list\">\n<li>Gender and age bias: tested each gender and each age group separately to remove any gender and age differences in the results<\/li>\n\n\n\n<li>&nbsp;Income: separated groups into high and low income groups so their nutrition styles and activity styles are somewhat similar.&nbsp;<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><span style=\"text-decoration: underline;\">Gender gaps in activity and obesity: <\/span>calculated 25th, 50th, and 75th percentile of daily steps for women and men in each country. They found that the gender gap explains 43% of the variation in inequality.&nbsp;<\/li>\n\n\n\n<li><span style=\"text-decoration: underline;\">City walkability analysis:<\/span> Walkability scores are based on nearby amenities and pedestrian friendliness. For the 69 U.S. cities with at least 200 users, they correlated walkability with city-level inequality.<\/li>\n\n\n\n<li><span style=\"text-decoration: underline;\">Impact of walkability on daily steps:<\/span> Linear regressions on walkability for subgroups for each age group, BMI categories, and gender from 24 cities with at least 25 users per subgroup shows that increase in walkability score showed increase in activity mainly in women.&nbsp;<\/li>\n\n\n\n<li><span style=\"text-decoration: underline;\">Simulating population-level changes in activity: <\/span>simulated two ways of spreading 100 extra steps per person per day:\n<ul class=\"wp-block-list\">\n<li><strong>Population-wide:<\/strong> Everyone gets 100 more steps.<\/li>\n\n\n\n<li><strong>Inequality-centric:<\/strong> The same total goes to the least active 5% to 9% of people.<\/li>\n\n\n\n<li>This approach reduced obesity prevalence by up to 8.3%, versus up to 2.3% for the population-wide approach. <\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><span style=\"text-decoration: underline;\">Data and Code availability: <\/span>They are linked to a github page on the project website.<\/li>\n<\/ul>\n\n\n\n<p class=\"has-foreground-color has-text-color has-link-color has-headings-font-family has-medium-large-font-size wp-elements-2 wp-block-paragraph\">Tools and Data Links:<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Tools: <a href=\"https:\/\/www.azumio.com\/apps\/argus\/overview\">https:\/\/www.azumio.com\/apps\/argus\/overview<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Data and Code: <a href=\"https:\/\/github.com\/timalthoff\/activityinequality.git\">https:\/\/github.com\/timalthoff\/activityinequality.git<\/a><\/p>\n\n\n\n<p class=\"has-foreground-color has-text-color has-link-color has-headings-font-family has-medium-large-font-size wp-elements-3 wp-block-paragraph\">Defining activity inequality:<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"901\" height=\"1024\" src=\"https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-content\/uploads\/2026\/11\/Screenshot-2026-11-08-at-5.25.19-PM-901x1024.png\" alt=\"\" class=\"wp-image-245\" srcset=\"https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-content\/uploads\/2026\/11\/Screenshot-2026-11-08-at-5.25.19-PM-901x1024.png 901w, https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-content\/uploads\/2026\/11\/Screenshot-2026-11-08-at-5.25.19-PM-264x300.png 264w, https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-content\/uploads\/2026\/11\/Screenshot-2026-11-08-at-5.25.19-PM-767x872.png 767w, https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-content\/uploads\/2026\/11\/Screenshot-2026-11-08-at-5.25.19-PM-1169x1329.png 1169w, https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-content\/uploads\/2026\/11\/Screenshot-2026-11-08-at-5.25.19-PM.png 1202w\" sizes=\"auto, (max-width: 901px) 100vw, 901px\" \/><\/figure>\n\n\n\n<p class=\"has-headings-font-family has-lead-font-size wp-block-paragraph\">The gini coefficient is a number between 0 and 1 which measures how unevenly something is shared across a group. 0 is associated with perfect equality, and 1 is associated with perfect inequality. It measures the distribution not the amount. For instance a rich country and poor country can have the same gini coefficient if the wealth spread is equally uneven. This gini coefficient comes from the Lorenzo curve. It measures the area between the diagonal and the curve, relative to the whole area under the diagonal. The more the curve bows, the larger the Gini. The Gini coefficient is the ratio of the area that lies between the line of equality and the Lorenz curve to the total area under the line of equality. The Gini coefficient is used to compute activity inequality. It is  commonly used to measure inequalities like financial inequality and the statistical spread. The Gini coefficient is based on the Lorenzo curve, which plots the share of the population\u2019s total average daily steps that is cumulatively recorded by the bottom X% of the population.The paper builds the Lorenzo curve from the tracked stepcount and computes the gini coefficient for each country. High gini coefficient means that the country has very few active people and low gini coefficient means that the activity spread is even. They found that the gap between men and women explains 43% of the variation in activity inequality across countries. A large gender gap pushes the Gini up, because the proportion is uneven as women are less active<strong>.<\/strong> They calculated the Gini for 69 U.S. cities and found that more walkable cities had lower activity inequality.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Overview of the Methods (Assessed through documentation): Tools and Data Links: Tools: https:\/\/www.azumio.com\/apps\/argus\/overview Data and Code: https:\/\/github.com\/timalthoff\/activityinequality.git Defining activity inequality: The gini coefficient is a number between 0 and 1 which measures how unevenly something is shared across a group. 0 is associated with perfect equality, and 1 is associated with perfect inequality. It measures [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-59","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-json\/wp\/v2\/pages\/59","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-json\/wp\/v2\/comments?post=59"}],"version-history":[{"count":7,"href":"https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-json\/wp\/v2\/pages\/59\/revisions"}],"predecessor-version":[{"id":272,"href":"https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-json\/wp\/v2\/pages\/59\/revisions\/272"}],"wp:attachment":[{"href":"https:\/\/classprojects.digitalwcu.org\/HMAI\/wp-json\/wp\/v2\/media?parent=59"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}