GBL Access
Team GBL

Team GBL

April 10th, 2020 5 min read

A-Z of Research

“We should totally get involved in the 36 Days of Type campaign on Instagram”, we all decided during our late-February Marketing team meeting. The remainder of our time slot was spent frantically trying to come up with economic terms related to the obscure letters, with xylophone steadily becoming more and more feasible as dehydration crept in and our patience dwindled.

a - asymmetric information is when one party in a transaction has more information than the other. In certain situations, information asymmetry obstructs beneficial transactions.

b - suppose you measure how one input affects the output. Bias is the amount by which you are wrong. This often happens when inputs are related to unseen things/beliefs.

c - that against which a measurement is made. A counterfactual is the answer to the question "Compared to what?"

d - deviation is the amount by which what you observe differs from any fixed reference point.

e - when something is endogenous, it is affected by something within the system under study. In many cases, you can tell whether something is endogenous if people can influence it by choice. When changes are endogenous, it is difficult to measure their effect.

f - when you don't have enough evidence to reject a hypothesis, you don't say "it's true", instead you "fail to reject" it. From a purely editorial perspective "fail to reject" amounts to a double negative, but sometimes research can be sassy.

g - generalizability is the extent to which research findings and conclusions conducted on a specific context can be applied to populations at large.

h - hypothesis is that what you assume before you start any investigation to find out whether it's true or not.

i - incentive is a benefit that motivates someone to do something.

j - job market is a market where employers search for employees and employees search for jobs. Far from a physical place, this concept defines the interplay between the players.

k - kurtosis is a measure of how outlier-prone a distribution is i.e. how much chances does the distribution have of having sizeable observations different from the average.

l - something that is lopsided is uneven because one side is lower or heavier than the other. If you say that a situation is lopsided, you mean that one element is much stronger, bigger, or more important than another element.

m - margin of Error is the amount by which you allow your claims to vary from reality. It is a safe bracket within which reality lies.

n - simply put, norm means the usual or in statistics it means the average.

o - outlier is that something or someone what/who is situated away or detached from the main body or system.

p - (statistical) power is the probability that a (statistical) test will detect differences when they truly exist. Think of (statistical) power as having the (statistical) "muscle" to be able to detect differences between the groups you are studying, or making sure you do not "miss" finding those differences.

q - you get a quasi-experiment when a noticeable change happens, but the researchers or those being affected by the change have nothing to do with it.

r - having no predictable pattern.

s - when information is conveyed from one party to another through (often) indirect means.

t - in research, testing is the act of gathering and processing information about individuals' ability, skill, understanding, or knowledge under controlled conditions.

u - the benefit people derive from consuming a good or service, or acting a certain way.

v - a characteristic that can differ from person to person, place to place, time to time, or situation to situation.

w - weight is an amount through which you can reflect the relative importance of an item amongst a group of items.

x - x variable is an example of a variable - usually the independent variable according to statistical norms (check our post on N). Being an independent variable implies x can take any value.

y - y variable is an example of another variable - usually the dependent variable according to statistical norms. Being a dependent variable implies y depends on x for its value.

z - zero sum games are games where one wins and by implication the other loses. In other words, one has to lose for the other to win.

These past 36 days have felt like 360 days, with the world being turned upside down by Covid-19. But even at home, even in such times, we realized there wasn't a dearth of ideas to do things fun. At GBL, that attitude resulted in the A-Z glossary you just read through, only to realize that at the end, we settled for x-axis… someone just had a breakup.

Related articles