Success in research: Difference between revisions
m Lumo moved page Success in research to Comparing amateur and professional research: Use a title that reflects the article question |
Explain what helps research succeed across independent, citizen science, and professional projects |
||
| Line 1: | Line 1: | ||
A | Success in [[Research|research]] starts with asking a question that evidence can answer. A useful result tells readers what was found, how it was checked, and where uncertainty remains. A project can succeed by documenting a little-known event, producing a reliable dataset, or discovering that a widely held explanation fits the evidence poorly. | ||
== Who does research? == | |||
An '''independent researcher''' pursues a project outside an institutional research appointment. They may work alone or with collaborators and may have years of training or practical experience. Someone who publishes local history videos, for example, may know the area's maps and archives well enough to notice that a frequently repeated story gives the wrong date for a flood. Sharing the relevant records lets a viewer check the claim. | |||
In '''citizen science''', members of the public take part voluntarily in scientific research. They may help choose the question, gather observations, interpret results, or lead a project. '''Professional researchers''' do research as part of their paid work. These roles can overlap. An independent researcher can contribute to a citizen science project, and a professional in one field can volunteer in another. [https://www.noaa.gov/office-education/citizen-science NOAA's introduction to citizen science] | |||
== What helps research succeed? == | |||
* '''Ask a question you can answer.''' If you want to date a flood, decide whether you mean when water first entered the streets or when it reached its highest point. Look for records that can distinguish the two. In a larger project, make sure everyone gathering observations uses the same definitions. | |||
* '''Leave a record of the evidence and decisions.''' Record where each observation came from and how you handled conflicting accounts. If you use a spreadsheet or code, keep the original data and explain the steps that produced your results. When working with others, agree on who will check the records and how contributors will be credited. | |||
* '''Check the result independently.''' Ask someone to trace a sample of claims back to the evidence. They may find that two newspaper stories drew on the same report, so both repeat the same date. Compare the accounts with a dated photograph or another source that could confirm or challenge them. Explain which parts remain uncertain. | |||
== What Galaxy Zoo shows == | |||
In the original Galaxy Zoo project, members of the public classified images from the Sloan Digital Sky Survey. The team combined multiple judgments and checked its results against classifications by professional astronomers for subsets of the galaxies. The two sets of classifications were broadly consistent for the simple galaxy shapes the project studied. Together, the contributors produced a catalog that other researchers could use. [https://arxiv.org/abs/0804.4483 Lintott and colleagues, Galaxy Zoo] | |||
A vote can help identify images that need another look. Reviewers should also check some of the images on which almost everyone agrees. If most people read the same blurry spiral galaxy as elliptical, high agreement can hide a shared error. | |||
== How could we compare different approaches? == | |||
When can a large group with little formal training produce more useful or original results than a small team of specialists? Start with a task you can measure. Ask volunteer, professional, and mixed teams to mark features in the same set of public images. Give them the same tools, instructions, and practice examples. Record relevant experience alongside formal education. | |||
Decide what counts as success before seeing the results. Count the features each team finds correctly, misses, or marks by mistake. Have reviewers check a sample against an independently prepared reference without knowing which team produced it. If the project is looking for novel findings, track which leads survive follow-up. It can take years to tell whether a result changes a field. | |||
Give teams equal work hours if you want to compare how they perform with the same amount of time. If you want to compare what each approach achieves within a budget, let team sizes vary and record training, coordination, and checking costs, along with unpaid hours. Repeat the study with different teams and images to see how far the result holds. | |||
For a classroom version, have students label a set of public images independently. Compare individual labels and the group vote with reference labels prepared ahead of time. Find one case where voting helped and one where the group would benefit from reviewing the image together. | |||
==Discussion questions, essay ideas, and learning related AI prompt ideas== | ==Discussion questions, essay ideas, and learning related AI prompt ideas== | ||
* What | * What could make an independently produced local history video more useful to a researcher than a published summary of the same event? | ||
* Write an essay | * Write an essay about how you would divide the work of studying a local stream among residents, independent researchers, and a laboratory team. Explain who would choose the question and how each group would check the others' findings. | ||
* Ask an AI tool to | * Ask an AI tool to design a fair comparison of volunteer and professional teams on a public dataset. Have it specify the outcome and what to do with ambiguous cases. Exchange plans and see whether two readers can score the same example consistently. | ||
==Readings from Wikipedia== | ==Readings from Wikipedia== | ||
| Line 38: | Line 43: | ||
== Research readings == | == Research readings == | ||
* [https://www.nationalacademies.org/read/25303/chapter/3 National Academies, Reproducibility and Replicability in Science]. The summary explains | * [https://www.nationalacademies.org/read/25303/chapter/3 National Academies, Reproducibility and Replicability in Science]. The summary explains how researchers check an original analysis and repeat a study with new data. | ||
== See also == | == See also == | ||