Probability vs. Non-Probability Sampling Approaches, Explained

UPI Study › Courses › Research Methods in Psychology › Free lessons › Lesson 8.3 Free with a free account

Probability vs. Non-Probability Sampling Approaches

Research Methods in Psychology · Chapter 8: Sampling and GeneralizationSection 8.3 · lesson 47 of 9411 min · free with a free account

← Linking Population, Sample, and GeneralizationNext lesson: Recognizing Common Sampling Challenges →

Earn 3 college credits for Research Methods in PsychologyRecommended for credit by NCCRS · Psychology 111 · self-paced · one payment, lifetime access
Enroll in Research Methods in Psychology · $250 →Download the syllabus (PDF)

In this lesson

What each family of methods can and cannot deliver.

  • Distinguish the two families on conceptual grounds
  • Identify the five major probability methods
  • Recognize the legitimate uses of non-probability methods
  • Match a sampling method to a research goal

Free study notes

Get this lesson's notes by email

The key points and 4 practice questions with answers, in your inbox. One email, free.

One email. Unsubscribe in one click.

Transcript

Everything said in the video, section by section (1,407 words). Read along, or search it.

Sampling and Generalization

Probability vs. Non-Probability Sampling Approaches

How you pick your participants decides what your numbers are allowed to mean. Every sampling method in social science falls into one of two families. This section maps both, and tells you what each one can deliver. You will distinguish probability from non-probability sampling on conceptual grounds. You will meet the five major probability methods, and when each fits. You will meet five non-probability methods, and their genuinely legitimate uses. And you will learn to match a method to a goal, which is the actual skill. One misconception gets corrected along the way, and it is a big one.

The Divide

Chance, or Something Else

Sampling methods divide into two broad families. In probability sampling, every member of the target population has a known chance of selection. Known, and greater than zero. The defining feature is that selection is governed by chance. Not by the researcher's judgment, and not by the participant's willingness. That matters because of what it unlocks. The laws of statistical inference apply, so confidence intervals and significance tests mean something. Statistical generalization to a defined population becomes possible. In non-probability sampling, selection probabilities are unknown, or vary in unmeasured ways. Judgment, convenience, or willingness decide who ends up in the sample. These methods cannot support statistical generalization in the formal sense. And yet the great majority of social-science studies use them.

Family One

The Five Probability Methods

Start with the probability family, which has five main members. Simple random sampling. Systematic sampling. Stratified sampling. Cluster sampling. And multi-stage sampling. All five share the same core commitment. Chance decides who is selected, and the odds are known in advance. What differs is how the chance is organized. Some methods treat the population as one undifferentiated list. Others carve it into subgroups first, for precision or for cost.

The Baseline

Simple Random and Systematic

Simple random sampling gives every member of the frame an equal chance. You number all the members, then select with a random-number generator. That is the conceptual gold standard, and everything else is measured against it. But it comes with a hard requirement. You need a complete sampling frame, which is often the limiting factor in real research. Systematic sampling is the practical cousin. You select every k-th member from an ordered list. And k is just the population size divided by the sample size you want. It is easier to implement, and it produces equivalent results. With one condition, and it is a real one. The list must have no periodic pattern that aligns with your sampling interval. If it does, you will systematically catch the same kind of case every time.

Structure and Geography

Stratified and Cluster

Stratified sampling adds structure before it adds randomness. You divide the population into subgroups called strata. By region, age, gender, or whatever category is relevant. Then you sample randomly within each stratum. That guarantees important subgroups appear at known rates. It also often produces more precise estimates than simple random sampling does. Proportionate stratification matches sample proportions to population proportions. Disproportionate stratification deliberately oversamples small but important groups. Cluster sampling solves a different problem, which is geography. You divide the population into naturally occurring groups, such as schools or hospitals. You randomly select some of those clusters, then study some or all of their members. That cuts logistical cost dramatically when your population is spread out. The price is somewhat less efficient estimation than simple or stratified sampling.

Combining Them

Multi-Stage Sampling, and Which to Use

Multi-stage sampling refuses to choose, and combines methods. Picture a national survey. It might first sample districts, which is cluster sampling. Then households within the sampled districts, at random. Then one individual per household, again at random. Multi-stage designs are common in large studies for a simple reason. They balance precision, cost, and feasibility, which no single method does well. So match method to circumstance. Simple random when you have a complete frame and no subgroup demands. Systematic when the frame is ordered and lacks periodic structure. Stratified when precision within subgroups matters. Cluster when the population is dispersed and individual sampling is impractical. And multi-stage for large national or international work.

Family Two

Convenience and Purposive

Now the non-probability family, starting with the two most common. Convenience sampling selects whoever is most readily available. Students in the researcher's own classroom. Shoppers passing a particular location, or volunteers from a social-media post. It is inexpensive and fast, and it provides no basis for statistical generalization. It is appropriate for piloting instruments and for exploratory work. It is not appropriate for estimating anything about a population. Purposive sampling is different in spirit, not just in procedure. The researcher selects participants by judgment about who can best inform the question. Variants include typical cases, extreme cases, and deviant cases. Or maximum-variation samples, chosen to span important dimensions. It is central to qualitative design, where depth and fit outrank statistical representation.

Family Two

Quota and Snowball

Quota sampling sets target numbers for subgroups, then recruits until they are filled. Fifty men and fifty women, say. Or twenty participants in each of three age categories. Structurally, that looks a great deal like stratified sampling. It differs in one critical way. Selection within each quota is non-random. The method is widely used in commercial polling and market research. But it produces biased estimates whenever the available people within a quota differ from the unavailable ones. Snowball sampling works by referral. You start with a few participants who fit the criteria, and ask them to refer others. It is valuable for populations that formal frames simply cannot identify. Migrants in irregular status. People with stigmatized conditions. Members of small subcultures. It is biased toward people who are well connected within that population. For hard-to-reach groups, it is sometimes the only practicable option there is.

Family Two

Self-Selection, and Where Each Belongs

The last non-probability method is self-selection sampling. You recruit whoever responds to an invitation. Online surveys, call-in radio polls, voluntary-response questionnaires. Such samples reflect the characteristics of those motivated to participate. And motivation almost always differs systematically from the broader population. So these samples are particularly vulnerable to bias, and should rarely be treated as representative. Still, every one of these five methods has a legitimate home. Convenience belongs in pilot work and instrument testing. Purposive belongs in qualitative research, selecting information-rich cases. Quota suits quick descriptive work where subgroup representation matters more than formal inference. Snowball reaches hidden populations no frame can list. And self-selection suits exploratory work, gathering opinions from those who want to share them.

A Correction

Non-Probability Is Not the Booby Prize

Now the correction promised at the start. It is tempting to treat non-probability sampling as a second-best option. Something you settle for when you cannot afford probability methods. That framing is misleading, and it produces bad decisions. For qualitative research, purposive sampling is not merely acceptable. It is actively preferable. Randomly selecting participants for an interview study would often be a methodological mistake. You would spend your limited interviews on people who cannot inform the question. So retire the question, is this probability sampling? Replace it with a better one. Is this method suited to the goal?

Fitting Method to Goal

Six Goals, Six Recommendations

Six common goals, and the approach each one calls for. Estimating prevalence in a population calls for probability sampling. Often stratified or multi-stage, depending on scale. Comparing well-defined subgroups calls for stratified probability sampling. With proportionate or disproportionate strata, depending on what you need. Studying a hidden or hard-to-reach population calls for snowball or other targeted methods. Building grounded theory from in-depth cases calls for theoretical or purposive sampling. Piloting a new instrument calls for convenience sampling. With explicit acknowledgment of its limits, which costs you nothing to state. And generating hypotheses calls for purposive or convenience sampling. Provided you have clear plans for follow-up testing afterwards. There is no universally best approach on that list. There is only fit.

Ethics

Who You Sample Is a Moral Choice

One more dimension, and it is not a technical one. Every sampling method carries ethical responsibilities. Probability sampling raises questions about how the lists were obtained. And whether contacting selected individuals is appropriate at all. Snowball sampling raises questions about social pressure. A referral from a friend is harder to refuse than a letter from a stranger. Convenience sampling can systematically exclude people without access to the recruitment channel. And quota sampling depends on category labels, often around gender, race, or ethnicity. Those labels require respectful and accurate handling. When findings inform policy, who was sampled determines whose interests the policy reflects. Systematically exclude the very poor, the very ill, the homeless, or the imprisoned. Exclude the linguistically isolated, or the digitally unconnected. And the policies built on that research may overlook precisely those people. Methodological choices have moral consequences. That is not a metaphor.

About Research Methods in Psychology at UPI Study

Research Methods in Psychology (Psychology 111) is a self-paced online course of 94 video lessons in 16 chapters, from the scientific method and research ethics to designs, data collection, sampling, measurement, validity, experiments, descriptive statistics and hypothesis testing. It carries a NCCRS credit recommendation for 3 semester credits.

Every lesson transcript is free on pages like this one, and the videos are free with a free UPI Study account. Enrolling adds a practice quiz after each lesson, the graded work and final exam that earn the credit, lifetime access, and an official transcript on completion. UPI Study has helped over 130,000 students with college credits. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.

Enroll now → Course details Download the syllabus All Psychology and social science courses Every course with PRO →

Compare: Research Methods vs Statistics

Why take it here

Why students take Research Methods in Psychology with UPI Study

How psychologists know what they knowResearch Methods in Psychology teaches how psychological studies are designed, run and judged: research ethics, surveys and observation, correlational and experimental designs, sampling, reliability and validity, and the statistics used to test hypotheses. It is a core course in psychology and many social-science degrees.
Credit at a fraction of the priceThe same 3 credits can cost $750 or more at a US college. Here it is one payment, self-paced, with lifetime access.
Finish your degree soonerTransfer credits reduce the courses you still owe, so you can graduate earlier or lighten a heavy semester.
Strengthen an applicationA completed college-level course on an official transcript shows admissions and scholarship committees you can do the work.
Study on your own scheduleNo start dates, no deadlines. Watch the lessons, practice, and take the final when you are ready.
130,000+students helped with college creditsACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.

Where the credit goes

Universities that work with NCCRS

Research Methods in Psychology carries a NCCRS credit recommendation. These universities work with NCCRS; some are UPI Study's signed partners. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.

Thomas Edison State UniversityThomas Edison State UniversitySigned partnerUniversity of the PeopleUniversity of the PeopleSigned partnerCharter Oak State CollegeCharter Oak State CollegeSigned partnerStrayer UniversityStrayer UniversitySigned partnerColumbia Southern UniversityColumbia Southern UniversitySigned partnerIndiana TechIndiana TechSigned partnerPenn StatePenn StateWorks with NCCRSSouthern New Hampshire UniversitySouthern New Hampshire UniversityWorks with NCCRSSUNYSUNYWorks with NCCRS +many more cooperating universitiesFind my college →

Find my college → Not listed? Ask Athena in the corner of this page, or book a 15-minute call.

FAQ

Questions students ask us

About Probability vs. Non-Probability Sampling Approaches

Is non-probability sampling worse than probability sampling?
Not necessarily; the better method is the one suited to the research goal. Probability sampling is needed to estimate prevalence or other population values, because only it supports formal statistical generalization. For qualitative research, purposive sampling is actively preferable, since randomly chosen interviewees may be unable to inform the question. Convenience samples suit pilot testing of new instruments, and snowball sampling can reach hidden populations that no sampling frame lists.
What is the difference between stratified and cluster sampling?
Stratified sampling divides the population into subgroups, such as region or age, and randomly samples within every stratum, guaranteeing each subgroup appears and often improving precision. Cluster sampling divides the population into naturally occurring groups, such as schools or hospitals, randomly selects some clusters, and studies members within them. Cluster sampling cuts costs for dispersed populations but usually gives less precise estimates.
How is quota sampling different from stratified sampling?
Both set target numbers for subgroups, but quota sampling selects people within each quota non-randomly, while stratified sampling selects them at random. A quota sample might recruit fifty men and fifty women from whoever is available until each quota fills. Quota sampling is widely used in commercial polling and market research, but it produces biased estimates when available people within a quota differ from unavailable ones.
What is snowball sampling and when is it used?
Snowball sampling starts with a few participants who meet the study criteria and asks them to refer others. It is used for hidden or hard-to-reach populations that formal sampling frames cannot identify, such as migrants with irregular status, people with stigmatized conditions, or members of small subcultures. Its main limitation is bias toward well-connected people, and referrals from friends can create social pressure to participate.

About the Research Methods in Psychology course

How many college credits is Research Methods in Psychology?
Research Methods in Psychology (Psychology 111) carries a NCCRS credit recommendation for 3 semester credits. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.
How much does Research Methods in Psychology cost and what is included?
$250, paid once, with lifetime access: all 94 video lessons, a practice quiz after each, the graded work and final exam, and an official transcript when you finish. Students who want several courses choose UPI Study PRO instead.
Which universities accept UPI Study credits?
UPI Study has signed partner agreements with universities such as Thomas Edison State University, University of the People and Charter Oak State College, and works with universities such as Penn State, SNHU and SUNY. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities. Use "Find my college" on this page to check yours.
How long does Research Methods in Psychology take?
It is self-paced with no deadlines. The 94 lessons total about 16 hours of video; most students add reading and practice time and finish in a few weeks to a few months, on their own schedule.
How are the exams done?
Quizzes and assignments are online in the course. The final exam is taken online with proctoring; a person on the UPI Study team checks assignments, usually within two days.
How do my credits reach my college?
When the course shows 100% complete, request the official transcript from your dashboard. It is sent to the college you name, and you can send it to more than one. Check your college first with "Find my college" on this page, or ask Athena.

Ready for the full course? Earn 3 credits in Research Methods in Psychology.

94 video lessons, a quiz after each one, and the graded work that earns the credit. Recommended for credit by NCCRS.

Enroll in Research Methods in Psychology →Download the syllabusEvery course with PRO →
Research Methods in Psychology
3 credits · NCCRSEnroll →