Student - MIR

[Student - MIR @MIRStudent22]

@IEGPAMIRSS Interesting! I think it would have a great impact on correspondence testing and its implementation, given that it would not allow researchers to use treatments such as "gendered names", which is standard practice to understand discriminatory hiring decisions.
#ieMethods

[Student - MIR @MIRStudent22]

Correspondence testing can have ethical concerns regarding employer entrapment, given that the focus of these experiments is to send out fake resumes to real job openings. That allows researchers to get data about common hiring practices, without recruiter's knowing.
#ieMethods

[Student - MIR @MIRStudent22]

The author further states that recruiters would most likely alter their behaviour if they knew they were being studied, and given that the researchers are only observing the practices (without prompting illegal behaviours), there is no major ethical concern.
#ieMethods

[Student - MIR @MIRStudent22]

However, in 2019 Eva Zschirnt, in her paper "Research Ethics in Correspondence Testing: An Update" argues that it is necessary to investigate these practices and study discriminatory practices in the labour force, regardless.

https://t.co/3F6y92mllj
#ieMethods

[Student - MIR @MIRStudent22]

@IEGPAMIRSS Very interesting study! I found it surprising that the researchers found evidence of selective perception bias, meaning that individuals' trust in their virtual teams is conditioned by their own expectations of trust. I wasn't aware of this type of bias in experiments.
#ieMethods

[Student - MIR @MIRStudent22]

@IEGPAMIRSS Definitely! I used this study as a foundation for my Research Design and used "gender" and "disability status" as treatments to further study the intersectionality and effect of these characteristics on employment opportunities.
#ieMethods

[Student - MIR @MIRStudent22]

I found Bertrand & Mullainathan's (2004). "Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination" very interesting. It is a great starting point to think about other correspondence tests.
https://t.co/feRgaurvBJ
#ieMethods

[Student - MIR @MIRStudent22]

@IEGPAMIRSS Very fair point. I would also add that self-reported behaviours or acts are often misreported since there isn't a clear guideline or checklist that respondents can follow. Furthermore, it is very dependent on subjective perceptions that can risk internal validity.
#ieMethods

[Student - MIR @MIRStudent22]

@IEGPAMIRSS However, this could be compared to similar experiments done in other regions, and then compare results. Also, it could be a good first step to take the experiment to the national and then regional level.
#ieMethods

[Student - MIR @MIRStudent22]

@IEGPAMIRSS Very interesting article. I definitely agree that it is too narrow and causes external validity issues. Given that Zürich is an incredibly wealthy area, the answers will not be able to be generalized to the broad population.
#ieMethods

[Student - MIR @MIRStudent22]

@IEGPAMIRSS That leads to validity threats. Internal, because the cultural factors might act as confounders and it is difficult to establish causal relationships. And external, because the results will change depending on the society in which the experiment takes place.
#ieMethods

[Student - MIR @MIRStudent22]

Falisse and Leszczynska (2020) study, using a lab-in-the-field design, the effects of anti-corruption messages on public service delivery. This novel design is (technically) supposed to increase internal and external validity.

#ieMethods
https://t.co/qR5afC7OKm

[Student - MIR @MIRStudent22]

@IEGPAMIRSS Definitely. I think that there is a cultural factor missing from this study, since taking a bribe could be considered a "tip" for providing a service, and in some countries, it is needed in order to do certain transactions.
#ieMethods

[Student - MIR @MIRStudent22]

The internal validity increases because of the researcher's control over the design, while external validity is increased by the implementation in the field. However, there are behavioural and cultural aspects that can affect the validity of this experiment.
#ieMethods

[Student - MIR @MIRStudent22]

Very interesting article studying the effect nationalism (provoked by qualifying for the World Cup) can have on interstate conflict. However, this study shows potential internal validity problems regarding endogeneity.
https://t.co/Hh1rggwy8V
#ieMethods

[Student - MIR @MIRStudent22]

Another explanation beyond that could be that if people are told their party is corrupt, they will think that since that is the norm anyways (we assume in this case), theirs at least will have connections to get them what they promise during electoral campaigns
#ieMethods

[Student - MIR @MIRStudent22]

Very interesting article by Arias et al. (2018). They conducted a field experiment in Mexico to study the effects that information on corruption would have on election outcomes, in which they collaborated with local NGOs.

https://t.co/6MreOeGCAa
#ieMethods

[Student - MIR @MIRStudent22]

They were interested in seeing if the voting outcomes would be altered by that information. However, they found it had little to no effect. That could be because of the fact that corruption is an ingrained cultural factor that is accepted by that society

#ieMethods

[Student - MIR @MIRStudent22]

They distributed pamphlets at random precincts (the units), which contained information about a particular party being corrupt. These were the treated units, while the other precincts did not receive any extra information and served as control units.

#ieMethods

[Student - MIR @MIRStudent22]

It is clear to see the methodology, especially as they describe the large number of attributes to be randomized (9 different characteristics), which is a defining feature of this type of experiment
#ieMethods

[Student - MIR @MIRStudent22]

This paper (Bansak, Kirk, Hainmueller, Jens and Hangartner, Dominik (2016)) clearly shows through the practical example of acceptance of asylum seekers in a given country, how a conjoint survey experiment works
https://t.co/oLqDPr6FLf
#ieMethods