Before collecting any data, you must consider both research ethics and researcher safety. Ethical considerations concern the rights and welfare of participants, including informed consent, privacy, confidentiality, and protection from harm. Researchers must ensure that relevant ethical guidelines are followed and that any required ethics approval has been obtained.
You must also consider your own safety when conducting fieldwork. Surveys, interviews, observations, and other forms of field research may require researchers to travel to unfamiliar areas, meet strangers, work alone, or enter locations where they may face physical, social, or political risks. Before beginning fieldwork, researchers should assess the location, time of day, method of travel, people they will be meeting, and any particular risks associated with the research topic or setting. Students should avoid unnecessary risks, inform someone where they will be, and withdraw from any situation in which they feel unsafe.
You should therefore plan data collection carefully before entering the field. This includes developing a structured data collection protocol explaining what will be collected, how it will be collected, where collection will take place, and how the resulting material will be recorded and stored. Good preparation helps ensure that data collection is ethical, safe, systematic, and consistent.
Archive and document research
Archive and document research uses existing materials, including government documents, diplomatic records, newspapers, speeches, websites, social media posts, photographs, and other digital materials. The process generally involves:
Develop a data collection protocol - specify what material will be collected, where it will come from, the period covered, selection criteria, and how it will be recorded and stored.
Identify sources - locate relevant archives, databases, collections, websites, or social media platforms.
Collect systematically - apply the same selection criteria and recording procedures throughout the research.
Record and store - keep full source details and organise materials using consistent filenames, folders, and reference-management software.
For websites and social media, remember that material can change or disappear. Relevant material should therefore be properly recorded and preserved where appropriate.
Observations
Observation involves systematically watching and recording behaviour, events, interactions, or environments. The process generally involves:
Develop a data collection protocol - specify what will be observed, where and when observation will occur, the researcher’s role, and how observations will be recorded.
Select the setting - identify the locations, events, or situations to be observed.
Conduct the observation - follow the protocol and record observations systematically.
Record and store - prepare detailed field notes and securely organise the collected material.
Reflect - consider how the researcher’s presence and interpretation may have influenced the evidence.
Preparing an observation protocol
Structured observation involves the systematic observation and documentation of a social setting. Before beginning, researchers should prepare an observation protocol: a detailed guide that identifies what should be observed and provides a consistent structure for taking field notes. The protocol helps ensure that the researcher records relevant information systematically rather than simply writing down whatever appears interesting at the time.
An observation protocol should begin with a basic record of each observation period, including the date, time, location, setting, participants, and duration. It should then provide space for different types of observations:
Direct observations - what did you actually see and hear?
Inferential observations - what do you think the observed behaviour or interaction might mean?
Analytic observations - how might the observation relate to your research question, concepts, or developing explanation?
Personal observations - what were your own impressions, reactions, or emotions?
Researchers should keep these categories separate wherever possible. In particular, distinguish what happened from your interpretation of what happened. This makes it easier to return to the original evidence during analysis.
After each observation period, update your notes as soon as possible. Add forgotten details or afterthoughts, record emerging ideas for analysis, and identify questions or issues to examine during future observations. The observation protocol should therefore develop into a systematic record of both what was observed and how your understanding of the research setting developed over time.
Surveys
Surveys collect standardized information from a sample of respondents and are particularly useful for examining attitudes, opinions, behaviours, and experiences. The process generally involves:
Develop a data collection protocol - specify the target population, sampling procedure, recruitment method, survey format, and procedures for collecting and storing responses.
Design the questionnaire - connect questions to the concepts being measured and use clear, neutral wording.
Pilot the survey - test the questionnaire with a small group and make necessary revisions.
Administer the survey - recruit the sample and collect responses according to the protocol.
Prepare and store the data - check, code, clean, and securely store responses for analysis.
Preparing a survey
A survey requires a survey protocol that explains how respondents will be recruited, how the questionnaire will be administered, and how responses will be recorded and stored. The protocol should also contain the questionnaire itself and establish consistent procedures so that respondents receive the same questions and instructions.
Two principles should guide survey design: do not confuse respondents, and consider every question from the respondent’s perspective. Researchers know much more about their topic than respondents and should not assume that respondents understand academic concepts, technical language, or the researcher’s intentions.
Questions should be simple, clear, and focused on one issue at a time. In particular:
Avoid jargon, slang, abbreviations, ambiguity, and vague language.
Avoid emotional language, prestige bias, and leading questions.
Avoid double-barrelled questions that ask about two things at once.
Do not ask questions respondents cannot reasonably answer, such as about high-level knowledge, or the distant future or past.
Avoid questions containing false or unsupported premises.
Ensure response categories do not overlap and provide a balanced range of possible answers.
Questions should be tested through a pilot survey before the main survey begins. This can reveal confusing wording, missing response options, technical problems, or questions that respondents interpret differently from what the researcher intended.
Different questions require different response formats. Common options include:
Open-ended - respondents answer in their own words.
Closed or partially closed-ended - respondents select from provided answers, sometimes with an “other” option.
Likert scale - measures levels of agreement or disagreement.
Quality - rates something from poor to excellent.
Quantitative judgement - asks respondents to estimate an amount or quantity.
Ranking - places several options in order.
Checklist - allows respondents to select all applicable options.
Classification - places respondents into categories.
Frequency - measures how often something occurs.
Importance - measures how important something is.
Likelihood - measures how likely an event or behaviour is.
Linear rating - asks respondents to place their answer along a numerical scale.
The response format should match what the researcher is trying to measure. Researchers should not automatically use the same type of scale for every question.
Researchers should also record what happens during recruitment. A low response rate can affect the quality and representativeness of the resulting data. Several stages can be monitored:
Location rate - percentage of the sampling frame successfully located.
Contact rate - percentage of located respondents successfully contacted.
Eligibility rate - percentage of contacted respondents who meet the study criteria.
Cooperation rate - percentage of eligible respondents who agree to participate.
Completion rate - percentage of participating respondents who complete the questionnaire.
Total response rate - percentage of the original sampling frame who ultimately complete the survey.
These figures help researchers identify where participants were lost during the survey process and assess whether non-response may have introduced bias into the final sample.
Interviews
Interviews collect detailed information about people’s experiences, knowledge, beliefs, interpretations, and decisions. The process generally involves:
Develop a data collection protocol - specify who will be interviewed, how participants will be recruited, where interviews will occur, how consent will be obtained, and how interviews will be recorded and stored.
Develop an interview guide - prepare questions connected to the research question while allowing appropriate follow-up questions.
Recruit participants - apply the sampling strategy and provide participants with the necessary information about the research.
Conduct the interviews - obtain consent and conduct each interview consistently with the agreed protocol.
Prepare and store the data - transcribe or organise notes, anonymise where necessary, and securely store the research material.
Preparing structured interviews
A structured interview uses a predetermined set of questions that are asked of each participant in the same order and, as far as possible, in the same way. It provides more consistency than semi-structured or unstructured interviewing and makes responses easier to compare across participants.
Researchers should prepare an interview protocol before beginning. This should contain the interview questions, their order, instructions for the interviewer, procedures for recording responses, and an introduction explaining the research and consent process.
Questions should be directly connected to the research question and the concepts being investigated. As with surveys, questions should be clear and understandable from the participant’s perspective. Researchers should:
use simple and neutral language;
avoid jargon, ambiguity, and unnecessarily complex questions;
ask one question at a time and avoid double-barrelled questions;
avoid leading or emotionally loaded questions;
avoid making assumptions about the participant’s knowledge or experience;
ensure participants can reasonably answer the question; and
arrange questions in a logical order, beginning with relatively straightforward questions before moving to more difficult or sensitive topics.
The interview should be piloted before formal data collection begins. This can identify questions that are unclear, repetitive, difficult to answer, or interpreted differently than intended.
Consistency is particularly important in structured interviews. The researcher should:
Introduce the research - explain the purpose and procedures and confirm consent.
Follow the interview protocol - ask each participant the same questions in the established order.
Remain neutral - avoid signalling approval, disagreement, or preferred answers.
Record responses consistently - use the same procedures for every participant.
Close the interview - provide an opportunity for questions and explain what will happen to the information collected.
Researchers should also record basic information about each interview, including the date, location or format, duration, and any circumstances that may have affected the interview.
Interviews may be audio-recorded with permission or documented through detailed notes. Recordings and transcripts should be labelled consistently and stored securely. Any identifying information should be managed according to the research protocol and ethical requirements.
Researchers should also keep brief field notes after each interview. These can record unusual circumstances, problems with particular questions, observations about the interview process, and ideas that may be useful during later analysis. This creates a record not only of participants’ responses but also of how the interview data were collected.
Preparing unstructured interviews
An unstructured interview is a flexible conversation in which the researcher explores a topic without following a fixed set of questions. Rather than asking every participant identical questions in the same order, the researcher allows the conversation to develop according to the participant’s experiences, knowledge, and responses.
Unstructured does not mean unprepared. Researchers should still develop an interview protocol before beginning. This should identify the purpose of the interview, the main topics to explore, procedures for consent and recording, and how interview materials will be stored.
Instead of preparing a detailed questionnaire, researchers develop a list of topics, themes, or broad opening questions. These should be connected to the research question while allowing participants considerable freedom to determine the direction of the conversation. Researchers should:
identify the main topics that need to be explored;
prepare broad, open-ended questions;
avoid leading or emotionally loaded questions;
use language appropriate to the participant;
prepare possible prompts for important topics;
consider how sensitive issues will be approached; and
pilot the interview approach where possible.
The researcher’s role is to guide rather than control the conversation. The researcher should:
Introduce the research - explain the purpose and procedures and confirm consent.
Begin broadly - use an open question that allows the participant to describe their experience or perspective.
Listen carefully - allow the participant to raise issues that the researcher may not have anticipated.
Probe important points - use follow-up questions such as asking for examples, explanations, or clarification.
Remain neutral - avoid signalling preferred answers or imposing the researcher’s interpretation.
Close the interview - check whether the participant wishes to add anything and explain what will happen to the information collected.
With permission, interviews should normally be recorded and later transcribed or carefully summarized. Researchers should also prepare field notes immediately after the interview, recording important observations, unexpected themes, problems, and ideas for future interviews.
Because each unstructured interview may develop differently, researchers must keep careful records of what was discussed, why particular topics were pursued, and how their interview approach developed over the course of the research. This flexibility is a major strength of unstructured interviewing, but it also requires careful documentation to maintain a systematic research process.
The guided academic paper - collecting data
You should now decide how you will collect the data for your guided academic paper. Your choice should follow from your research question, measurement decisions, and sampling strategy.
First, identify the most appropriate method: archive and document research, observation, surveys, interviews, or a combination of these. Then develop a structured data collection protocol explaining exactly how you will collect and record your evidence. Ensure that all ethical guidelines are followed before beginning data collection. Record:
the data collection method you will use and why;
the sources, settings, or participants from which data will be collected;
your data collection protocol;
any survey questions, interview questions, or observation framework you will use;
how the data will be recorded, organised, and stored; and
any practical or ethical problems you anticipate.
Your protocol should be detailed enough that another researcher could understand how you intend to collect your data and could follow the same basic procedure.
Support an academic - Buy me a Coffee