Drug purchasing offline and online (social media and messaging apps) in Poland – online survey data (2023–2025)

Abstract

General Information Funding Information The dataset is a joint output of two research projects that developed and administered the questionnaire together: “Rhizomatic networks, circulation of meanings and contents, and offline contexts of online drug trade” (2022–2025; PLN 956 620), funded by the Polish National Science Centre (Narodowe Centrum Nauki, NCN), call OPUS 22, grant no. 2021/43/B/HS6/00710; Principal Investigator: Piotr Siuda (Kazimierz Wielki University in Bydgoszcz, Poland). “Badanie e-rynków narkotykowych w Polsce – handel z wykorzystaniem platform społecznościowych” (“A study of online drug markets in Poland – trade via social media platforms”), a three-year research project ([2023–2025]) carried out at the Institute of Law Studies of the Polish Academy of Sciences (Instytut Nauk Prawnych PAN, INP PAN), project leader: Adam Stasiak. The project was a public health task co-financed by the Gambling Problem Solving Fund (Fundusz Rozwiązywania Problemów Hazardowych) and commissioned by the National Centre for Addiction Prevention (Krajowe Centrum Przeciwdziałania Uzależnieniom, KCPU) (under agreement no.275/KT/HN/DBMWM/2023). Data Collection Context Data Source The data come from an anonymous online questionnaire (CAWI) administered in Polish on the LimeSurvey platform. The link to the questionnaire was disseminated primarily through social media and in cooperation with Polish non-governmental organizations working in the field of drug policy and harm reduction – Wolne Konopie, Polska Sieć Polityki Narkotykowej (Polish Drug Policy Network) and Mestosław – which were asked to advertise the survey and to encourage their followers to complete it. Responses were collected between 30 December 2023 and 5 June 2025, in two main waves (December 2023–January 2024 and July–August 2024) followed by a small number of late responses. The sample is a non-probability, self-selected sample of Polish-speaking internet users, skewed towards young adults and towards people who follow drug-policy and harm-reduction communities online. It is not representative of the Polish population or of people who use drugs in Poland. In the INP PAN/KCPU project the survey was complemented by five questions placed in a nationwide, representative Ariadna omnibus survey of adult internet users (N = 1,117; 1–3 June 2025), which served to contextualize the survey results; the Ariadna data are not part of this deposit. Purpose The questionnaire was designed to describe the demand side of digital drug markets in Poland, i.e., the practices, motivations and perceived safety of people who buy psychoactive substances, and to compare purchasing via the internet (social media platforms, messaging apps, websites and darknet marketplaces) with purchasing through face-to-face contacts. The NCN project studies internet behavior related to the online drug trade in Poland and seeks to (1) investigate how the open internet, including social media, is used in the drug trade; (2) outline the significance of darknet platforms in the distribution of drugs; and (3) explore the exchange of content related to the drug trade between the surface web and the darknet, along with the meanings constructed within the drug subculture. The survey provides the user perspective that complements the project’s platform data. The INP PAN/KCPU project explored the realities of drug e-markets on Facebook, Instagram and Telegram – identifying the markets, describing trends and changes in online drug trading, and characterizing market participants on both the demand and the supply side. It combined three components: (1) a quantitative analysis of sales offers posted on messaging platforms, (2) the present survey of people buying drugs online, supplemented by the representative Ariadna omnibus, and (3) in-depth interviews with people who supply drugs online. The final, optional item of the questionnaire invited respondents to leave contact details in order to take part in these interviews. The contact details are not part of the deposited dataset. Within this design the survey was intended to: document which substances are bought online and offline and how often; identify the platforms, sources of information, payment and delivery methods used in online purchases; capture protective strategies (e.g. escrow, legit checks, secret chats) and the criteria used when choosing a seller; measure attitudes towards online purchasing (safety, quality, availability, price); record experiences of fraud in both settings; and describe the transition from buying to social supply and to selling for profit Data Content Data Description The dataset is a single data file with 1,733 records, one per questionnaire response, and 159 variables. It contains complete and partially completed responses: 1,559 respondents reached the last (seventh) page of the questionnaire, 140 stopped on page 6 and 34 on page 5; the variable lastpage records the last page reached, and questions on pages not reached are empty. Of the 1,733 respondents, 1,571 declared that they had ever bought psychoactive substances and 162 that they had not. Respondents who had never bought (p0 = 2) were routed past all subsequent modules, including the demographic questions, which is why their demographic variables are empty; the base for the demographic variables is the 1,571 buyers. Respondent characteristics (among those who reached the demographic module): gender (n = 1,561) – 937 men, 527 women, 35 non-binary persons, 62 preferred not to say; country of residence (n = 1,558) – 1,445 Poland, 113 outside Poland; size of place of residence (n = 1,558) – 216 village, 138 town up to 20,000 inhabitants, 181 up to 50,000, 154 up to 100,000, 287 up to 500,000, 582 city over 500,000. Year of birth was entered as a free number (n = 1,468); most respondents were born between 1990 and 2007, i.e. the sample is dominated by people aged 18–30 at the time of the survey, with a small share of respondents aged 15–18. The questionnaire (about 30 main questions, many of them multiple-response, plus quantity items and open “other – please specify” fields) is organized into the following modules: Screening and purchase channels: whether the respondent has ever bought psychoactive substances and whether face to face, via the internet, or both (p0, p1, p1_1, p1_2). Purchases in the real world: sources (close friends, acquaintances, dealers known only for that purpose, random persons); substances ever bought and substances bought in the last 3 months (a list of 15 categories: marijuana, hashish, cannabis oil, cocaine, amphetamine, methamphetamine, MDMA, ketamine, 3-CMC/mephedrone, psilocybin mushrooms, GBL, LSD, prescription drugs, other, don't know); quantities bought in the last 3 months for eight substances (p2–p6). Purchases via the internet: substances ever bought and bought in the last 3 months (same list) and quantities (p7–p9, p8b–p8h); platforms used (Facebook, Instagram, Telegram, Snapchat, websites, other); how the respondent learned about buying drugs online (friends, strangers, own search, platform recommendation/chance, other); payment methods (cash, BLIK, bank transfer, barter, cryptocurrencies, other); delivery methods (delivery by seller/courier, taxi/Uber/Bolt, self-collection, parcel locker, courier shipment, other); safety practices (escrow, legit check, secret chat, PrivNote, other, none); criteria for choosing a seller (p10–p15). Attitudes: four five-point agreement items comparing online with face-to-face purchasing in terms of safety, quality, ease of access and price (p16–p19). Fraud: experiences of being cheated when buying face to face and online and the form of the fraud (p20a–p20bbb). Social supply and selling: giving substances to friends free of charge and passing them on for profit, and to whom (p21–p22a). Demographics: year of birth, gender, country of residence, size of place of residence (p25–p28). Open comments on buying drugs online or in the real world (p23) and “other – please specify” text fields attached to the closed questions. Variable names follow the questionnaire numbering (p0–p28) with suffixes for individual response options (e.g., p3_1–p3_15) and open “other” fields (e.g. p3a). Variable labels and value labels are stored in Polish in the data file; an English translation of every variable and value label is provided in the accompanying codebook. Data Cleaning, Processing, and Anonymization The responses were exported from the survey system to an SPSS data file. The survey did not collect names or IP addresses, and no such data are present in the deposited file. The following steps were taken to protect respondents: All date and time information was removed. The original export contained the response timestamps startdate and datestamp (accurate to the second); both variables were deleted in their entirety, as was the submission date. The optional closing item of the questionnaire, in which respondents willing to take part in follow-up interviews could leave contact details, was removed in its entirety together with all answers (variable p99 in the original export; 87 respondents had left e-mail addresses, telephone numbers or messenger handles). All fourteen open-text variables (p3a, p5a, p7a, p10a, p10b, p11a, p12a, p13aa, p13bb, p14a, p15a, p20aaa, p20bbb, p23) were screened for direct and indirect identifiers using automated pattern matching followed by manual review of every non-empty answer (1,038 answers in total). Twelve identifiers found in twelve answers given by ten respondents – usernames/handles and addresses of specific websites and shops – were removed and replaced with the markers [REMOVED: username] and [REMOVED: URL]. The markers [REMOVED: e-mail] and [REMOVED: phone number] are documented in the codebook but do not occur in the deposited data. Generic references to the technology (e.g. “strony .onion”) and the names of publicly known marketplaces were left unchanged, as they do not identify individuals and are analytically relevant. The technical variable id is the respondent identifier assigned by the survey system and cannot be linked to any person. The variable lastpage records the last questionnaire page reached. No other cleaning was applied. Numeric free entries – year of birth (p25) and quantities bought in the last 3 months (p6a–p6h, p9a, p8b–p8h) – are stored exactly as entered and contain implausible values (e.g. birth years such as 0, 420 or 20010425; quantities such as 9999). Users should apply their own plausibility rules and treat such entries as missing. The unit of the quantity items is grams. Coding conventions of the multiple-response items: for p1_1 and p1_2 the values are 1 = selected and 0 = not selected; for all other multiple-response items only the value 1 = selected is stored, and an empty cell means either “not selected” or “question not asked.” To distinguish the two, use the routing variables (p1, p4, p8, p20a, p20b, p22), as documented in the codebook. The Polish labels of p20bb_1–p20bb_3 read “w świecie rzeczywistym” (“in the real world”) although the items refer to online purchases and are asked only of respondents who answered p20b = 1 (this labeling error does not affect the data or analyses); the online quantity items are named p9a (marijuana) and p8b–p8h (the remaining substances), rather than following a single sequence; in question 10 the option “Snapchat” is stored as p10_6 and “Other” as p10_5. File Formats and Variables/Fields The dataset consists of the following files: drug_purchasing_survey_PL_2023-2025.sav – SPSS System File (UTF-8, compressed) with 1,733 cases and 159 variables, including Polish variable labels and value labels. drug_purchasing_survey_PL_2023-2025.csv – the same data as a CSV file for users without SPSS. UTF-8 with byte-order mark, comma-separated, numeric codes (value labels are given in the codebook), empty cell = missing. drug_purchasing_survey_PL_2023-2025_codebook.csv – codebook listing, for each of the 159 variables: the variable name, the question text as stored in the file (Polish), an English translation, variable type and SPSS format, value labels in Polish and English, the routing/filter condition, the number of valid entries and notes. UTF-8 with byte-order mark. drug_purchasing_survey_PL_2023-2025_questionnaire.pdf — the original Polish questionnaire. Single-response items are coded 1 = yes, 2 = no (3 = “I do not want to answer” where offered); p1 uses 1 = only in the real world, 2 = real world and internet, 3 = only internet (the value label 998 = not applicable is defined in the file but does not occur in the data; respondents who never bought have p1 empty); the attitude items p16–p19 use a five-point scale from 1 = completely agree to 5 = completely disagree; p11, p26, p27 and p28 are nominal variables whose categories are listed in the codebook. Missing data are system-missing (empty) throughout. Ethical Considerations Ethics Statement Participation in the survey was voluntary and anonymous. The opening page of the questionnaire informed respondents about the aims of the study, the institutions conducting it, the anonymity of responses and the possibility of stopping at any point; the most sensitive questions offered an explicit “I do not want to answer” option. The survey did not ask for any information that would allow the identification of respondents; the only personal data collected – contact details voluntarily left by respondents willing to take part in follow-up interviews – were stored separately, used solely to arrange the interviews and are not included in the deposited data. Ethical approval was granted by the Research Ethics Committee of the Institute of Law Studies of the Polish Academy of Sciences (Komisja ds. Etyki Badań, Instytut Nauk Prawnych PAN). A set of data-handling policies aimed at ensuring safety and ethics in dark web and online drug market research, followed in both projects, has been outlined in: Harviainen, J.T., Haasio, A., Ruokolainen, T., Hassan, L., Siuda, P., Hamari, J. (2021). Information Protection in Dark Web Drug Markets Research [in:] Proceedings of the 54th Hawaii International Conference on System Sciences, HICSS 2021, Grand Hyatt Kauai, Hawaii, USA, 4–8 January 2021, (ed.) Tung X. Bui, Honolulu, HI, pp. 4673–4680. The ethical safeguards of the INP PAN/KCPU project (pseudonymization, removal of names of persons, profiles, groups and channels, modification of details that could allow the recognition of individual cases) were utilized. Because respondents self-report behavior that is illegal under Polish law, the dataset was prepared so that no direct identifiers remain and open-text answers were screened for indirect identifiers. Users of the data are asked to use it for research purposes only, not to attempt to re-identify respondents or to combine the data with other sources for that purpose, and to treat the open-text answers as sensitive material that should not be quoted in a way that could point to an individual.

Topics & conditions

Detected automatically by Psychedelic Commons from the title and abstract, not labels from the source. They may be incomplete.

Keywords