an

adopt

want

necessarily

The ball is now in your court: think how to use skills developed through the book to formulate good research questions, form collaborative teams, and shape the future of scholarship.

several concern has

whereas the latter may be even hampered by such approaches

knowledge gaps and interest gaps

.

and articulations against established scholarship and ability to transform those ideas into computational code

both sociological imagination and technology imaginations are required to identify interesting research questions

in

Therefore, we see hybrid approaches were humans and computational analysis are deeply engaged in the sense-making process as one emergent approach for doing computational social science work.

There is a range of opportunities is

A valid question is that how to choose the concepts focused on.

those

'

studies analysis

it often academics build on previous traditions and scholarship.

These highlights invite us to consider new ways of doing research, relaxing the traditional strong division between theory-first (deductive) and theory-last (inductive) thinking.

(With any and all meanings of theory - disciplines within social sciences and computational social science may have different ideas about what constitutes theory.)

loan

.

Underlying are deeper differences between what is the role of theory in research.

Computational methods are still seeking to understand how it should be done to ensure clarity of concepts in its work

discuss about the relevant conceptual thinking and therefore, lacks a theoretic contribution

theoretic

solving

Put more bluntly: if someone plans to shoot oneself in the foot, computational methods allow using bazooka instead a smaller gun.

add references

This was a long introduction to say that this book tries not to advocate for a particular way of doing research, but embrace openess on these questions.

[width=.7]figures/research_process

In computational social sciences, the operationalisation process involves transforming a theorethical perspective into an algorithm.

What are the three different aspects of how computational social science challenges existing social science research practices?

Technological imagination refers to the ability to use computational tools to capture the interesting questions.

.

Based on my own experiences, these challenges can even become emotionally charged. Differences in core assumptions may be difficult to discuss even among academics. However, for computational social science to succeed, everyone must be able to cooperate across these types of collaborations. We will return to the opportunities and problems of multidisciplinary collaboration in Chapter 11.

we

(

Table 1.1: Different approaches of computational social science

However, the gist of the work is to ensure that the research and findings speak with existing social science literature.

rigorous

component

Within this book, social science theory refers to the collective knowledge developed and conceptualised in social scientists over decades.

None of them is wrong. Instead, they highlight different pieces of the puzzle.

decreases a negative social phenomenon or increases a positive social phenomenon.

Based on the symposium series, most people have their main discipline related to computers of information technology.

Rather, it seems to adapt methods familiar to physicists and complexity scientists to social science research problems.

paradigm of conducting social science research.

check for a good quote?

After the tournament, the winning strategy was tit-for-tat, a simple approach where one defected only in cases where the opposite side had already defected in the previous turn.

tool

complex

What — if any — differences are there between advanced quantitative research methods and computational research methods?

the core idea is not the novel data but advanced and more rigorous methods that allow closer investigation than before.

In this book, I will advocate that this is not the case.

predicting variables from the data.

Thus, their work shows that political polarisation occurs in political blogs. There were clear divisions based on parties.

problems

:

refs

Therefore, it is hard to justify that computational social science would be a holistic discipline but rather a multidisciplinary mesh of scholars doing research using computation with social science questions in mind.

including me

makes all social scientists

Most intriguing but also most controversial thoughts have asked

or

.

a larger number than in the original hallmark study.

digital society

drive

.

contacted to the researchers

approached

At

Tastes, Ties, or Time

Figure 4.3: Schematic of the experimental design from Schultz et al. (2007). The field experiment involved visiting about 300 households in San Marcos, California five times over an eight-week period. On each visit, the researchers manually took a reading from the house’s power meter. On two of the visits, they placed doorhangers on each house providing some information about the household’s energy usage. The research question was how the content of these messages would impact energy use.

in order do

Figure 3.7: Demographics of respondents in W. Wang et al. (2015). Because respondents were recruited from XBox, they were more likely to be young and more likely to be male, relative to voters in the 2012 election. Adapted from W. Wang et al. (2015), figure 1.

statistical statistical

Wikipedia.

???

¿sec:algorthmically-confounded?

[@king_how_2016

Figure 2.2:

custom-made

American Association of Cancer Researchers

impartial

dominate

Academics of Science

recoding

Athey (2017), Cederman and Weidmann (2017), Hofman, Sharma, and Watts (2017), (???), and Yarkoni and Westfall (2017)

contests

effort building

they

about about

a project to classify a million galaxy classification

Write an email summarizing what you think is happening and recommend a course of action.

it is

Design advice

difference-of-differences

difference-of-means

difference-of-means

difference-of-differences

research

excludibility

these these

have have

pre-treatment

Olmstead

other other

are who

use

the response propensity

XBox

XBox

XBox

XBox

XBox

particular

treatment assignment

to treatment

donates

inaccessible,

Yi(1)

1

have have

need

lab

employment

are who

embedded

[Y(1,Wi(1))−Y(0,Wi(0))]

think

[, ] In a lovely paper, Lewis and Rao (2015) vividly illustrate a fundamental statistical limitation of even massive experiments. The paper—which originally had the provocative title “On the Near-impossibility of Measuring the Returns to Advertising”—shows how difficult it is to measure the return on investment of online ads, even with digital experiments involving millions of customers. More generally, the paper clearly shows that it is hard to estimate small treatment effect amidst noisy outcome data. Or stated diffently, the paper shows that estimated treatment effects will have large confidence intervals when the impact-to-standard-deviation (δ¯yσ

[, , ] Michel et al. (2011) constructed a corpus emerging from Google’s effort to digitize books. Using the first version of the corpus, which was published in 2009 and contained over 5 million digitized books, the authors analyzed word usage frequency to investigate linguistic changes and cultural trends. Soon the Google Books Corpus became a popular data source for researchers, and a 2nd version of the database was released in 2012. However, Pechenick, Danforth, and Dodds (2015) warned that researchers need to fully characterize the sampling process of the corpus before using it for drawing broad conclusions. The main issue is that the corpus is library-like, containing one of each book. As a result, an individual, prolific author is able to noticeably insert new phrases into the Google Books lexicon. Moreover, scientific texts constitute an increasingly substantive portion of the corpus throughout the 1900s. In addition, by comparing two versions of the English Fiction datasets, Pechenick et al. found evidence that insufficient filtering was used in producing the first version. All of the data needed for activity is available here: http://storage.googleapis.com/books/ngrams/books/datasetsv2.html In Michel et al.’s original paper (2011), they used the 1st version of the English data set, plotted the frequency of usage of the years “1880”, “1912” and “1973”, and concluded that “we are forgetting our past faster with each passing year” (Fig. 3A, Michel et al.). Replicate the same plot using 1) 1st version of the corpus, English dataset (same as Fig. 3A, Michel et al.) Now replicate the same plot with the 1st version, English fiction dataset. Now replicate the same plot with the 2nd version of the corpus, English dataset. Finally, replicate the same plot with the 2nd version, English fiction dataset. Describe the differences and similarities between these four plots. Do you agree with Michel et al.’s original interpretation of the observed trend? (Hint: c) and d) should be the same as Figure 16 in Pechenick et al.) Now that you have replicated this one finding using different Google Books corpora, choose another linguistic change or cultural phenomena presented in Michel et al.’s original paper. Do you agree with their interpretation in light of the limitations presented in Pechenick et al.? To make your argument stronger, try replicate the same graph using different versions of data set as above.

see Appendix Table 10

evaluting

evaluting

δ¯yσ

impact-to-standard-deviation (δ¯yσ

diffently

2nd version of the corpus, English dataset

1st version of the corpus, English dataset

“1880”, “1912” and “1973”

Social research is a process of asking and answering questions about human behavior.

web-based experiment

DRAFT Please do not distribute