2021 Kaggle DS & ML Survey
¡ñ Methodology and survey flow logic
Survey Methodology: 2 Survey Flow Logic: 3
Copyright By PowCoder代写 加微信 powcoder
Survey Methodology:
¡ñ The 2021 Kaggle DS & ML Survey received 25,973 usable responses from participants in 171 different countries and territories.
¡ñ You can find the full list of questions and answer choices in the file “kaggle_survey_2021_answer_choices.pdf”.
¡ñ Responses to multiple choice questions (only a single choice can be selected) were recorded in individual columns. Responses to multiple selection questions (multiple choices can be selected) were split into multiple columns (with one column per answer choice).
¡ñ To ensure response quality, we excluded respondents that were flagged by our survey system as ¡°Spam¡± or “Duplicate. We also dropped responses from respondents that spent less than 2 minutes completing the survey, as well as responses from respondents that selected fewer than 15 answer choices in total.
¡ñ To protect the respondents¡¯ privacy, free-form text responses were not included in the public survey dataset, and the order of the rows was shuffled (responses are not displayed in chronological order). Likewise, if a country or territory received less than 50 respondents, we grouped them into a group named ¡°Other¡± for the sake of anonymity.
¡ñ An invitation to participate in the survey was sent to the entire Kaggle community (anyone opted-in to the Kaggle Email List). The survey was also promoted on the Kaggle website (via both banners and popups) as well as on the Kaggle Twitter channel.
¡ñ The survey was live from 09/01/2021 to 10/04/2021. We allowed respondents to complete the survey at any time during that window.
¡ñ The survey data was released under a CC 2.0 license:
https://creativecommons.org/licenses/by/2.0/
Survey Flow Logic:
¡ñ The full list of questions and answer choices can be found in the file: kaggle_survey_2021_answer_choices.pdf. The file contains footnotes that describe which questions were asked to which respondents. Additional details are described below.
¡ñ Respondents with the most experience were asked the most questions. For example, students and unemployed persons were not asked questions about their employer. Likewise, respondents that do not write code were not asked questions about writing code.
¡ñ Follow-up questions were only asked to respondents that answered the setup question affirmatively.
¡ð Question 18 (which specific ML methods) was only asked to respondents that selected the relevant answer choices for Question 17 (which categories of algorithms).
¡ð Question 19 (which specific ML methods) was only asked to respondents that selected the relevant answer choices for Question 17 (which categories of algorithms).
¡ð Question 28 (which specific product) was only asked to respondents that selected more than one choice for Question 27-A (which of the following products).
¡ð Question 29-A (which specific AWS/Azure/GCP products) was only asked to respondents that selected the relevant answer choices for Question 27-A (which of the following companies).
¡ð Question 30-A (which specific AWS/Azure/GCP products) was only asked to respondents that selected the relevant answer choices for Question 27-A (which of the following companies).
¡ð Question 33 (which specific product) was only asked to respondents that selected more than one choice for Question 32-A (which of the following products).
¡ð Question 35 (which specific product) was only asked to respondents that selected more than one choice for Question 34-A (which of the following products).
¡ð Question 37-A (which specific product) was only asked to respondents that answered affirmatively to Question 36-A (which of the following categories of products).
¡ñ For questions about cloud computing products, students and respondents that have never spent money in the cloud were given an alternate set of questions that asked them ¡°what products they would like to become familiar with¡± instead of asking them ¡°which products they use most often¡±. For questions with alternative phrasing, the questions were kept separate, and question types were labeled with either an ¡°A¡± or a ¡°B¡± (e.g. Q29A, Q29B, … , Q37A, Q37B).
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com