Skip to main content

Family Recipes

It is Lent right now, and even though I am not religious, I still decided to try and curb one of my addiction at this time. I my case, it was sweets, and I have to admit, that first days were harder, than I imagined. I guess there was some addiction there, like I would have joked for the last couple of years.

I think this is why I went through the recipes, that I had collected. I realized that way too many of them were sweets. Not counting sweets, there was only one recipe, that I still wanted to try. Another thing, that I could get rid of. (This is always a time for celebration :) )

I also found some of the recipes, that I got from my two grandmothers. Since I might regret not having them in the future, I will write them down here. Even if they are sweets.

What I always assumed where my grandmother cookies (but then I realized, that she did not put chocolate in them):

Ingredients:

  • 200g of flour (polnozrnata pšenična moka)
  • 0,5 little spoon of baking soda (jedila soda)
  • 100g of butter
  • 75g of sugar (trsnega sladkorja)
  • 75g of chocolate
  • 1 egg
  • 1 little spoon of vanilla (vanilijev extrakt)

Maybe grandmother cookies (?):

  • 2 dl of water
  • 70 dag of butter
  • pinch of salt
  • 100g of flour
  • 3 eggs

Bake for 20 minutes at 200°C

Some chocolate cookie things that my other grandmother once got:

  • 3 eggs (whole)
  • 15 dag of sugar
  • 10 dag of butter
  • 32 dag of flour
  • a little of soda

  • walnuts, raisins, chocolate, (soncn? - maybe sunflower seeds? I can't read)

So I am putting it there, mostly so I make it easier to myself for throwing it away.

The Fear of Boredom

Do you know these conversations, where the conversation itself is a everyday thing, but they when you really think about it, you realize why there are problems in your life? Well, I recently had a conversation like that with a friend of mine. And there were three points in that conversation, that eventually led me to realize what is holding me back (and how useless it is):

  1. I told him that I am afraid of boredom
  2. He observed, that I know, how to shrug other ideas off
  3. I told him, that this is because I had too many of them

I mean, there are like everyday normal thing, that come up in the conversation all the time. Well, expect the boredom part, that one was a surprise even for me.

So things started to change very quickly after that. For one thing, the first thing I did was delete a lot of my files in the Later folder. This is a folder, where I keep all the things that I am currently working on: the articles that I want to read, the code that I am writing, the collection of ideas that I had and so on. A lot of it went away.

This is also an example of action coming before the insight. Because this is the first thing that I did, before I realized the rest of it.

Then I looked back, and I that insight came: I had a couple of interesting opportunities, that I did not follow up on, because I did not want to add something new to the plate. I did not know, how to handle another project on top of what I already had. So, I let the opportunities slip, which could make my life a lot more interesting.

But on the other hand, I am an INTP. I might not have the big vision stuff going on, but I can spot many interesting ideas in a single day. I mean, every time I come from the Python Meetup, I have more ideas that I could implement in a month, without having a conversation with a single purpose. I get way to many ideas from books or lectures and so on. That is before people offer me a chance to work on something.

And then I realized, that out of fear of not having enough opportunities (something my mind and the world already showed me that I do), I am sabotaging myself in taking them.

That made me realize, that I better finish the things that I started (my two master thesis, the working version of UExperience - though I know this will me a lot more long term as far as support goes, but by then, it will be a lot smaller demand on my time), eliminate what I don't want to do (I am looking at you Introduction to Cognitive Science 2) or things that I had procrastinated way too long on (like that analysis project of Arrowverse, which had from deep learning to graph analysis in it, learning lips and so on).

And I could eliminate a lot of internet communication. I am sorry, but I already have a hobbies that I can do, when I don't have energy for something else, and this is drawing and watching my favorite series in other languages (currently German). It is still going to be there, in case I ever need some more input, right?

I finally got to the end of this insight (about a week and a half after the conversation - it felt longer), when I was rereading the book Focus by Leo Babauta. Which talks about how to find some focus in our lives. And then the association when to the book Deep Work by Cal Newport and then to the So Good they can't Ignore You also by Cal Newport. And these three books were sort of a social proof, that I am on the right track. :)

I think this has been a lot easier for me, because I had dabbled in minimalism for years. So eliminating clutter is something that had become easier through time. Maybe for somebody else, the road would be a bit harder. And it is also interesting to see, if this road to minimalism will lead to bigger abundance.

But I think that even more important is, that we all need some sort of a mirror to see, what we are doing. So that in fear of something, we don't end up doing something, that will make it more likely to come true. Or, as a proverb, that I remember from one of the fanfiction stories: "One often meets one’s fate on the road one takes to avoid it."

I need to learn to post blogs when I write them as well - this one was posted more than a week after writting it

Learning Statistics with Basketball

As I am reading some of the statistical articles, to help me with the master thesis, I had come across one, that might be interested to more people.

This short (4 pages) article talks about how high-school students could be more motivated in statistics, but comparing Michael Jordan and LeBron James. So it is interesting to people who had interest in basketball (there are some analysis and data there), education and statistics.

The article is Understanding summary statistics and graphical techniques to compare Michael Jordan versus LeBron James

Social Skills and Math Skills

library(tidyverse)

I had to tutor a lot of people in mathematical or statistical part of economics and in programming. And a lot of times (though far from always), there were these moment, when I could see that people were not getting it, but I was not sure why.

I remeber a case in my master studies, where one of my classmates from abroad had problems, and wanted my help. So I said yes. Well, the problem was, that I eventually found out, that he did not even uderstand some of the basic mathematical principles. Like using the letters to stand for the numbers and linear regression. How does one explains this? I don't know, because to be, it was clear from the start.

It was only when I started tutoring in the programming, that I was able to formulate some hypotesis. One interesting things, that I had seen, with the people I was tutoring, was the use of the theory of mind. At least some of their problems steamed from their expectation, that the computer has a theory of mind and that this is a conversations.

I mean, it is, but it is a lot more structured. It is a conversation, where everything has to be explicitly said or was at one point explicitly agreed on, and these agreements can be checked. There is no reading the intention going one.

Maybe this is the reason, why some people are afraid of the code? Because the code/computer might take offense?

So, my hypotesis was, that this social understanding would impeed the ability to do math and code.

I found the article titled The Empathizing-Systemizing Theory, Social Abilities, and Mathematical Achievement in Children, that used the systemizing-empthatizing to try and research this, in out case, the empthazing would be connected to the empathy/social skills and so on. The people lower on empthaizing were better at math (calculating at this level). Systemizing was not connected in this stage.

But by the time people come to the university, systemizing was conneted to math intelligence, as discivered in article Systemisers are better at maths. Subject and gender differences in math dissapread when controled for systemizing.

This is as far as ability goes. In the article Testing the Empathizing–Systemizing theory of sex differences and the Extreme Male Brain theory of autism in half a million people, one of the analysis that they did was the difference in systemizing and empathizing between STEM and non-STEM employees. The STEM employes were lower in empthaizing (beta = -1.10) and higher in systemizing (beta = +1.27).

Now, if anybody is interested in the short, but very readable analysis of gender differences in STEM, I suggest the report Why don’t more girls choose to pursue a science career?

So, as far as the literature goes, there is some indication, that mathematical interest or skill could be connected to social skills.

So in order to get one more piece of the information, I will try to find some data and do analysis in this direction as well.

Country Level Analysis of Agreeablness and PISA Mathematical Scores

While the effect on the country levels can be different than the effects on the individual levels, it can still be useful to check it. Expecially, since a lot more country-level data is already avalable.

First I combined two dataset. The personality came for the article The Geographic Distribution of the Big Five Personality Traits: Patterns and Profiles of Human Self-Description Across 56 Nations, which can be found here: https://www.toddkshackelford.com/downloads/Schmitt-JCCP-2007.pdf . And the mathematical data came for the 2015 PISA results, that can be found here: http://pisadataexplorer.oecd.org . I combined the dataset, so it only included the countries, that were represented in both (40 countries).

data_country = read.csv("data/data-personality-math.csv")
head(data_country)
Country Math E A C N O
Argentina 409 49.10 42.75 48.18 55.05 50.83
Australia 494 48.98 47.51 45.87 50.82 50.07
Austria 497 50.61 45.90 46.73 49.69 49.29
Belgium 507 45.99 45.07 43.03 53.60 54.59
Brazil 377 45.89 45.86 45.38 53.14 49.16
Canada 516 48.32 49.14 49.05 50.58 48.75

Now, the social skills (that I am interested in) is most connected to agreeablness. So this is what I am going to be using for this analysis.

Let us first see the distribution of the two variables, that will be used.

ggplot(data_country, aes(x=Math)) + 
       geom_histogram(bins=20) +
       xlab("PISA Avreage Score Levels") + 
       ggtitle("Distribution of Mathematical Skills in Dataset")

Histogram of Mathematical Skills

ggplot(data_country, aes(x=A)) + 
       geom_histogram(bins=20) +
       xlab("Avreage Agreeablness Level") + 
       ggtitle("Distribution of Agreeablness in Dataset")

Histogram of Agreeablness Levels

And now for the connection.

ggplot(data_country, aes(x=Math, y=A)) + 
       geom_jitter(width=0.01, alpha=0.3) +
       ylab("Avreage Agreeablness Level") + 
       xlab("PISA Mathematical Level") + 
       ggtitle("PISA Math Skills and Agreeablness")

Scatterplot of Math Skill and Agreeablness Level

PISA results are standardized to 500 mean and 100 standard deviation over all the countries. The personality took US as standard and had all the countries standardized with US having the mean of 50 and standard deviation of 10. Since the scaling can be potential problem in statistical analysis, I will devide PISA scores with 10. This way, that will both be standardized on the similar scale (but not the same).

data_country$Math <- data_country$Math/10
head(data_country)
Country Math E A C N O
Argentina 40.9 49.10 42.75 48.18 55.05 50.83
Australia 49.4 48.98 47.51 45.87 50.82 50.07
Austria 49.7 50.61 45.90 46.73 49.69 49.29
Belgium 50.7 45.99 45.07 43.03 53.60 54.59
Brazil 37.7 45.89 45.86 45.38 53.14 49.16
Canada 51.6 48.32 49.14 49.05 50.58 48.75

Now, I will build a linear regression model, where I will try to predict mathematical scores from agreeablness.

model_country = lm(data_country$Math ~ data_country$A)

While the coeficient is -0.24 (so in the direction, that I would predict), I did not have a high enough sample to make it statistically significant. So this might just be a chance of random fluctuation.

summary(model_country)
Call:
lm(formula = data_country$Math ~ data_country$A)

Residuals:
   Min     1Q Median     3Q    Max 
-9.595 -2.489  1.317  3.140  6.736

Coefficients:
               Estimate Std. Error t value Pr(>|t|)    
(Intercept)     58.4208    12.7601   4.578 4.91e-05 ***
data_country$A  -0.2426     0.2708  -0.896    0.376    
---
Signif. codes:  0 *** 0.001 ** 0.01 * 0.05 . 0.1   1

Residual standard error: 4.596 on 38 degrees of freedom
Multiple R-squared:  0.02069,   Adjusted R-squared:  -0.005086 
F-statistic: 0.8026 on 1 and 38 DF,  p-value: 0.3759
ggplot(data_country, aes(x=Math, y=A)) + 
       geom_jitter(width=0.01, alpha=0.3) +
       geom_smooth(method=lm, color="red") + 
       ylab("Agreeablness") + 
       xlab("Math Skills") + 
       ggtitle("Connection Between Agreeablness and Math")

Graph of Connection Between Agreeablness and Math Skills

Just for my own interest, I want to see, which one would have the highest effect amoung personalty dimentions. I will again use linear regression.

model_country_5 = lm(data_country$Math ~ data_country$A + data_country$C + data_country$O + data_country$E + data_country$N)
summary(model_country_5)
Call:
lm(formula = data_country$Math ~ data_country$A + data_country$C + 
    data_country$O + data_country$E + data_country$N)

Residuals:
    Min      1Q  Median      3Q     Max 
-10.053  -2.184   1.636   2.686   5.458

Coefficients:
               Estimate Std. Error t value Pr(>|t|)   
(Intercept)    151.4562    54.6188   2.773  0.00895 **
data_country$A  -0.0740     0.3350  -0.221  0.82649   
data_country$C  -0.5721     0.3204  -1.786  0.08306 . 
data_country$O  -0.3000     0.2818  -1.065  0.29452   
data_country$E  -0.4551     0.6164  -0.738  0.46542   
data_country$N  -0.7286     0.4199  -1.735  0.09177 . 
---
Signif. codes:  0 *** 0.001 ** 0.01 * 0.05 . 0.1   1

Residual standard error: 4.42 on 34 degrees of freedom
Multiple R-squared:  0.1896,    Adjusted R-squared:  0.07048 
F-statistic: 1.591 on 5 and 34 DF,  p-value: 0.189

It seems, that when all the personality dimentions are added in, the agreeablness have the least affect. Instread it seems that neuroticism and maybe openess are much more interesting in this regard.

Disturbing Technology Design

Here I just wanted to share an article with you all. The article Motivating Learners by Nurturing Animal Companions: My-Pet and Our-Pet (actually a scientific conference presentation, but who cares) talks about, how the Tamagochi-es were used to make students more motivated for learning.

I mean, there are ethical problems already - the first of all the assumption that more studying (and cooperation) is good, which my itself is moral evaluation to the people having a worse emotional landscape because if it. Yes, I am sure of this, since they included the quote in the article, which was something like "Id the pet is sad, then so am I". Don't forget, these are children in school, not adults.

People anthropomorphise machine (just check, what happened to people talking to ELIZA), but we generally don't make people worse off, because somebody else did not do something. Except in action movies, where this is done by the bad guys. But for the people with very strong anthropomorphising, I don't think there is much of a difference. So I am left wondering, did nobody find this manipulative?

(Well, except my professor from Vienna, who mentioned this as a lack of ethical standard in human-computer interaction research until recently)