Showing posts with label demographics. Show all posts
Showing posts with label demographics. Show all posts

Sunday, June 7, 2020

COVID-19 and CSI without Dead Bodies

Since March 14, thirteen out of the last 14 posts on the blog have been on the Coronavirus pandemic.  One was on the new hate group numbers from the Southern Poverty law center.  With states starting to reopen after the I thought I would take a look at how the pandemic has affected traffic on this site according to Google Analytics.  I compared the traffic from March 14 to today (June 7) to the previous period (Dec 19,2019 to March 13, 2020).

Overall, the site has received 20.4% more traffic for the current period to the previous (1,133 to 941).  The average number of page views per session increased by 18.23% (1.47 pages/session for this period to 1.24 pages/session for the last period).  However, the average time spent on a page was 30.5% lower for the current period (0:26) compared to the previous (0:38).

Although the overall number of users on this site was up by 20.4%, the number of users from the U.S. was down by 12.46% (752 to 859).  There was number of users was up from China (49 to 2), India (19 to 3), the U.K. (18 to 10), Hong Kong (15 to 0), Germany (13 to 3), Japan (11 to 3) Mexico (11 to 5) and Canada (8 to 5).  Overall the number of users for this site from outside the U.S. was higher for this site with 381 for this period to 82 from outside the U.S.  This gives an increase of 365% from outside the U.S.

The ratio of new users to returning users for this period was 93.7% (1,103) new to 6.3% (74) returning for this period to 89.4% (897) to 10.6% (106) for the previous.  For those users that have their gender identified by Google the ratio of males to females was 51.01% (177) females to 48.99% (170) males for this period.  For the previous period it was 29.89% (55) for females versus 70.11% (129) for males.

For those users whose age is identified by google, the largest increase in users was in the 35-44 age group with a 180% increase (70 to 25) followed by the 65+ age group at 173.68% (52 to 19), and the 25-34 age group at 126.67% (102 to 45).  There were increases for all the age groups.  

The results here appear to be mixed.  There are more users for this site for the pandemic posts.  According to the graph in the image at the top of the post, the largest spike in number of users between the period occurred in the week from May 17 to May 23.  The post for this period was the one on COVID-19 testing in zip-codes in Johnstown.  The traffic for this post was almost exclusively from the U.S.  More specifically, it was mostly from Johnstown and the surrounding areas.  Without this post, the gap in users between the US and the rest of the world would be even greater.

**Related Posts**

Tuesday, June 6, 2017

Educational Attainment and the % Uninsured Explain Trump's % of the Vote with Philly Considered

I know it's been a while since I've posted last.  I've been busy with running for city council and other personal issues.  In my last big data post, I looked at county level simple correlations with education, uninsured, child abuse, and Trump's % of the vote.  This post will look at multiple regression for county level variables.  This is to sort out which effects are strongest when adjusting for the the others.

When adding the variables education (defined as % of population with a bachelor's degree or higher), % uninsured, and % change in child abuse cases with Trump % of the vote gives the following model with Philadelphia County included in the data set.  The coefficients are the values for the regression equation:

Trump % = 105.86 - 1.31* (% Bachelors) - 1.85*(% Uninsured) 

both variables are statistically significant (with a p-value of less than 0.05) and the model accounts for 62% of the variability in Trump's % of the vote.  Of this the % with a bachelor's degree accounts for 58% of the variability.

Standard Error
t Stat
Lower 95%
Upper 95%
Bachelors %
Uninsured %

The scatterplot below shows the actual values for each county (in blue) and the predicted values (in orange) for the above equation.  For example, Forest county had an actual % of the vote for Trump of 70.1% but the model predicts 79.3% for Trump with a predictor value of 8.3% uninsured and 8.5% with a bachelors degree or higher.

The coefficient for the % uninsured is significant if Philadelphia county is included but not if it is not.  The graph below shows that the association is weaker for this relationship accounting for only 4% of the variability.  Both charts show that Philadelphia county is considered. 

Philadelphia county is an outlier for the counties in Pennsylvania in much the same way as Washington, DC is for the other states. It had a big influence on the uninsured data so it was included in this model.  Both Philadelphia and DC are demographically different from the surrounding counties and states respectively.  The % of the white population in the counties and the states may be an important explanatory variable for the model. 

**Related Posts** 

Concentration of Hate Groups Predict Hate Crimes (if you consider DC) and Trump Vote (if you don't)


Hate Groups and Trump's Vote%: Predictive effect present when education and poverty are considered


More Hate Groups in States Where Trump and Clinton Win (and in DC Where He Lost)

Tuesday, May 24, 2016

Clinton 7 Times More Likely to Win in States Where Trump Wins

Clinton Win
Sanders win
Trump Win              
Trump Lose             

In my post from two weeks ago, I noticed that there were higher levels of hate group activity in states where Trump and Clinton won their primaries.  Does that mean that members of hate groups are more likely to support Clinton and Trump?  Not necessarily. 

The table at the top of this post shows the states won by Clinton and Sanders where the Republicans have also had their contests.  Fisher's Exact test shows that there is a significant association between states where Clinton wins and state where Trump wins.  The odds ratio of 7.37 says that Clinton is more than 7 times more likely to win in states where Trump wins than in states where he lost.

Fisher's Exact Test for Count Data

data:  z
p-value = 0.007189
alternative hypothesis: true odds ratio is not equal to 1
95 percent confidence interval:
  1.451117 51.760487
sample estimates:
odds ratio 
As a follow up to my post from two weeks ago I also looked at the mean concentration of hate groups in the four groups of states (Trump win-Clinton win, Trump win-Sanders Win,
Trump lose-Clinton win, and Trump lose-Sanders Win).  The means are charted above.  The chart suggests greater hate group concentration in states where Trump & Clinton win with the lowest concentration being in states where Trump loses and Sanders wins.  It also suggests somewhat higher levels activity in states where Clinton wins as opposed to the states where Sanders wins.  The face that the lines are not parallel in the chart above suggests an additive effect of Trump and Clinton winning.  These effects were not significant in a two-way analysis of variance due to grossly unequal numbers of states contributing to each of the four means.

The "state" with the highest concentration of hate groups, DC at 26.8 groups/million residents, was not included because the Democrats won't have their contest there until June 14.  Trump placed third there.

The higher concentration of hate groups in states where Clinton and Trump win may reflect support from these groups on Trumps side while fear of these groups could be driving Clinton's support on her side.  This is borne out in the exit polls from these states with nonwhite voters overwhelmingly supporting Clinton in these states and with prominent white supremacists supporting Trump nationwide.

**Related Posts**

More Hate Groups in States Where Trump and Clinton Win (and in DC Where He Lost)

SPLC Hate Group Update: Washington, DC has an Increase in Activity

What Would a Trump Presidency Look Like?