Most BS Criminology theses in the Philippines use a descriptive-correlational or descriptive-comparative design. The reason is practical rather than theoretical: the respondents available to a criminology student — PNP personnel, BJMP officers, barangay officials, residents of a barangay — are reachable by survey and rarely reachable by anything else. Your design should follow the verb in your statement of the problem and the access you can actually secure.
What are the designs a criminology thesis realistically chooses between?
Five, in rough order of how often Philippine criminology programs approve them.
Descriptive-survey. Measures the level, extent or status of something in one group. “Level of awareness of Republic Act 9165 among senior high school students in X.” No relationships, no comparisons. Simplest to defend, easiest to criticise as thin.
Descriptive-correlational. Measures two or more variables and tests whether they are related. “Relationship between community policing practices and residents’ perceived safety.” The most common criminology design in the country.
Descriptive-comparative. Measures one variable across groups and tests for differences. “Comparison of job stress among police personnel when grouped by length of service, rank and station assignment.”
Qualitative — phenomenological or case study. Explores lived experience through in-depth interviews. Suits topics where numbers cannot reach: the reintegration experience of former persons deprived of liberty, the experience of first responders at a crime scene.
Mixed methods, usually explanatory sequential. A survey first, then interviews with selected respondents to explain the numbers. Strong, and roughly double the work — take it only if your timeline genuinely allows.
How do I tell which one my study needs?
Read the verb in each numbered sub-problem and match it.
- “Determine the level of…” or “describe the extent of…” — descriptive-survey.
- “Determine whether there is a significant relationship between…” — descriptive-correlational.
- “Determine whether there is a significant difference when grouped according to…” — descriptive-comparative.
- “Explore the experiences of…” or “describe the lived experience of…” — qualitative.
- Both a relationship sub-problem and an experience sub-problem — mixed methods, or cut one of them.
Where studies get into trouble is having a quantitative title and a qualitative sub-problem hiding at number five, or the reverse. If the design named in your Chapter 3 does not answer every sub-problem in your Chapter 1, the panel will find the one it does not answer. The broader decision between the two paradigms, and what each commits you to, is worked through in the comparison of quantitative versus qualitative research for a thesis.
Why is descriptive-correlational so dominant in criminology?
Because criminology questions are naturally about relationships between things you can only observe, not manipulate. You cannot randomly assign barangays to receive more patrols. You cannot experimentally vary a police officer’s length of service. What you can do is measure both variables as they stand and test whether they move together.
The honest limit is that correlation is not causation, and criminology panels ask about it more than most because policy recommendations are usually the point of the study. Write your Chapter 5 recommendations as “association suggests” rather than “leads to,” and put the limitation in Chapter 1 rather than waiting for the panel to supply it.
Can a criminology thesis use an experimental design?
Rarely at undergraduate level, and almost never with human subjects in a justice setting. Quasi-experimental designs do appear — a pre-test and post-test around a crime prevention seminar or a drug awareness programme is a legitimate one-group design, with the obvious weakness that anything else happening during the interval competes as an explanation.
True experiments with control groups are ethically and practically closed to you in this field. Do not propose one to look ambitious.
What decides my design more than theory does?
Access. Three constraints settle criminology designs before methodology ever gets a say.
Institutional permission takes weeks. Surveying PNP personnel normally means a request routed through the Chief of Police and often the Provincial or Regional Office. BJMP respondents require clearance from the jail warden and usually the regional director. Barangay respondents need a courtesy letter to the Punong Barangay. Start the letters the week your title is approved.
Persons deprived of liberty are a protected population. Any study with PDL respondents is a full ethics review, not a courtesy letter — a captive population cannot give ordinary voluntary consent, and your school’s ethics committee will treat it accordingly. Many programs quietly discourage it for undergraduate theses because the approval alone can outlast the semester.
Crime data is not as available as you think. PNP crime statistics are not published in the granular, downloadable form students assume. Requesting station-level figures is its own approval process. A design that depends on obtaining records is a design that can fail through no fault of your methodology.
Choose the design your access supports. A well-executed descriptive-correlational study beats a mixed-methods design that stalls at the jail gate in October.
What goes in the research design section of Chapter 3?
Three things, in three or four sentences. Name the design. Define it, with a citation to a methods author. Justify it against your specific sub-problems.
This study employed a descriptive-correlational research design. A descriptive-correlational design describes variables as they naturally occur and determines the degree of relationship between them without manipulation (Creswell & Creswell, 2018). The design was appropriate because the study sought to determine the level of community policing practices and residents’ perceived safety, as stated in sub-problems 2 and 3, and to establish whether a significant relationship exists between them, as stated in sub-problem 4.
Notice the third sentence names the sub-problem numbers. That single habit answers a defense question before it is asked. The rest of the chapter — locale, respondents, sampling, instrument, data gathering procedure and statistical treatment — follows the sequence set out in the section-by-section guide to writing Chapter 3 research methodology.
Which sampling technique goes with which design?
Descriptive-correlational and descriptive-comparative studies with a known population — the residents of a barangay, the personnel of a police station — normally use simple random or stratified random sampling, with the sample size computed by Slovin’s formula at a 5% margin of error. Stratify when your comparison sub-problem groups respondents by rank, station or barangay, because simple random sampling can leave a stratum with four respondents.
Where the population is unknown or unlistable, purposive sampling with stated inclusion criteria is acceptable, provided you state the criteria explicitly and acknowledge the limitation on generalisability.
Qualitative designs use purposive sampling to saturation. Phenomenological studies in Philippine criminology programs typically run somewhere between eight and fifteen participants, with the actual number justified by data saturation rather than fixed in advance.
Does the design determine my statistical treatment?
Yes, almost entirely. A descriptive-survey design needs frequency, percentage, weighted mean and standard deviation. A descriptive-correlational design adds Pearson r, or Spearman rho when the data is ordinal or fails normality. A descriptive-comparative design adds a t-test for two groups or one-way ANOVA for three or more, with the non-parametric counterparts as the fallback.
If your design and your listed tests do not match, one of them is wrong. What each test is for and how to phrase its justification is covered in the explainer on the statistical treatment of data in a thesis, and the test-by-test selection logic works the same way it does in the comparison of statistical tests for accountancy studies.
What will the panel ask about my design?
Four questions, near-universally. Why this design and not another. What its limitations are. Why you did not use an experimental design. And how you can recommend a policy change from correlational evidence.
Prepare a two-sentence answer to each and rehearse them out loud. The full range of what Philippine panels ask, organised by chapter, is set out in the guide to the questions asked during a thesis defense.
Getting Chapter 3 written once the design is settled
Choosing the design is your judgement call. Writing the methodology chapter around it — the design justification tied to sub-problem numbers, the sampling computation, the instrument description, the statistical treatment entries — is formatting work, and it is where criminology groups burn the fortnight before the proposal defense.
Tesify drafts Chapter 3 from your own statement of the problem, keeps the design, the sampling and the statistical treatment consistent with each other, and formats your citations in APA 7th as you write. Start your thesis on Tesify and get the chapter drafted while your permission letters are still in transit.
Frequently asked questions
What is the most common research design in criminology theses in the Philippines?
Descriptive-correlational, followed by descriptive-comparative. Both fit survey data from reachable respondent groups such as police personnel, barangay officials and residents, and both answer the relationship and difference sub-problems that criminology titles typically promise.
Can a criminology thesis be purely qualitative?
Yes, and phenomenological designs suit topics about lived experience that numbers cannot reach. Check your program’s research manual first — some Philippine criminology departments require a quantitative component at undergraduate level, and that requirement overrides methodological preference.
How many respondents does a criminology thesis need?
For a quantitative study with a known population, compute the sample with Slovin’s formula at a 5% margin of error and report the computation in Chapter 3. For a phenomenological study, eight to fifteen participants is typical, with the final number justified by data saturation rather than set in advance.
Do I need permission to survey police officers?
Yes. Requests are normally routed through the Chief of Police for a station and through the Provincial or Regional Office for anything larger, with your adviser’s endorsement and a letter from your dean attached. Allow several weeks and begin as soon as your title is approved.
Can I interview persons deprived of liberty for my thesis?
Only with full ethics committee approval plus BJMP or BuCor clearance, because a captive population cannot give ordinary voluntary consent. The approval process is long enough that many undergraduate programs discourage it. Confirm with your adviser and ethics committee before committing your title to it.
Is a descriptive design too simple for a thesis?
A purely descriptive design can be criticised as thin if it only reports levels and stops. Adding a relationship or difference sub-problem, which turns it into a descriptive-correlational or descriptive-comparative study, is usually enough to satisfy a panel without adding significant work.
Where do I get Philippine crime statistics for Chapter 1?
PNP directorate reports and Philippine Statistics Authority compilations are the usual starting points, but neither publishes station-level data in downloadable form. Expect to file a formal request for local figures, and do not design a study that depends on records you have not yet been granted.
Can I change my research design after the proposal defense?
Usually only with your adviser’s approval and often a written justification to the panel, since a design change ripples through your sampling, instrument and statistical treatment. If access to a respondent group falls through, raise it immediately rather than collecting from a different group and adjusting the chapter afterwards.
What is the difference between research design and research method?
The design is the overall plan for answering your questions — descriptive-correlational, phenomenological, quasi-experimental. The method is how you collect the data within that plan — a survey questionnaire, an interview guide, document analysis. Chapter 3 names the design first, then describes the method.