TL;DR: The independent variable is the presumed cause, predictor or grouping factor; the dependent variable is the outcome you measure. Rewrite your own title as "the effect of A on B" or "the relationship between A and B" and the two fall out immediately. Beyond those, you need control variables (held constant), extraneous ones (acknowledged as limitations) and, if they apply, moderating and mediating ones. The step that decides whether your study is actually doable is the operational definition, and the step that decides which statistical test you may run is the level of measurement. Fix both before collecting anything.
Variables look like the easiest part of research methodology and cause a surprising share of failed proposals. The reason is that the label is not the problem; the measurement is. A student who can correctly point at the independent variable but cannot say in one sentence how it will be measured has not finished the job.
DepEd's Practical Research 2 curriculum guide places this right at the start of the subject, alongside understanding the nature of quantitative research: the competency is "differentiates kinds of variables and their uses" (CS_RS12-Ia-c-3). It comes first because everything downstream, the statement of the problem, the instrument, the statistical treatment, is built on it.
Which subject is this for? Practical Research 1 and 2 are applied subjects of the 2016 K to 12 Senior High School curriculum, and that structure is being replaced. DepEd Order No. 017, s. 2026 institutionalises the Strengthened Senior High School Curriculum for Grade 11 from SY 2026-2027 and Grade 12 from SY 2027-2028, and its Annex A lists Research 1 and Research 2 as 80-hour Academic Track electives in the Field Experience cluster rather than applied subjects everyone takes. Grade 12 learners not enrolled in pilot schools continue under the 2016 curriculum for SY 2026-2027. What a variable is does not change. See our guide to the Strengthened SHS Curriculum.
What Counts as a Variable
A variable is any characteristic that takes different values across your units of analysis. If it does not vary within your study, it is not a variable in that study, it is a constant.
This trips people up. If you survey only Grade 11 students, grade level is not a variable in your research; it is a fixed feature of your population. If you survey Grades 11 and 12, it becomes one. The same characteristic can be a variable in one study and a constant in another, and it depends entirely on who you sampled.
Independent vs Dependent
| Independent variable (IV) | Dependent variable (DV) | |
|---|---|---|
| Also called | Predictor, explanatory, grouping, treatment variable | Outcome, criterion, response variable |
| What it does | Presumed to cause, predict or distinguish | Presumed to respond or be predicted |
| In a title | The "A" in "effect of A on B" | The "B" |
| In an experiment | What the researcher manipulates | What the researcher measures |
| In a survey | What already differs between respondents | What you measure to see whether it tracks the IV |
The test that works on your own title. Rewrite the title in one of two frames. If "the effect of A on B" reads correctly, A is independent and B is dependent. If only "the relationship between A and B" reads correctly, you have a correlational study, and the labels are conventional rather than causal.
That distinction is not pedantry. In a correlational design, calling one variable "independent" does not license a causal conclusion. Our guide to writing a research hypothesis covers how to word the conclusion so it stays within what the design supports.
The Other Kinds, and What Each Is For
| Kind | What it is | What you do about it |
|---|---|---|
| Control | A variable you deliberately hold constant so it cannot vary | State it in Chapter 3, for example "all respondents from one school" |
| Extraneous | Anything outside your study that could affect the DV | Acknowledge in limitations; control what you reasonably can |
| Confounding | An extraneous variable related to both IV and DV, so it can fake a relationship | Control it, measure it, or declare it as a serious limitation |
| Moderating | Changes the strength or direction of the IV-DV relationship | Include only if a sub-problem tests it |
| Mediating | Sits between IV and DV and explains how the effect happens | Include only if a sub-problem tests it |
| Intervening | A broader term often used interchangeably with mediating in Philippine manuals | Follow your department's usage |
Most senior high school and undergraduate theses need only the first three rows. Moderators and mediators require additional statistical machinery and additional sub-problems, and adding them to a title without testing them is a common way to promise more than the study delivers.
Worked Example 1: A Correlational Title
The studies in this section are illustrative, constructed for this guide. They are not real research.
Illustrative title: "Daily Smartphone Screen Time and Self-Reported Sleep Quality Among Grade 11 Students in a Public Senior High School in Cebu City."
- Independent variable: daily smartphone screen time.
- Dependent variable: self-reported sleep quality.
- Control variables: grade level and school, both held constant because only Grade 11 students in one school are surveyed.
- Extraneous variables to declare: caffeine intake, part-time employment, household sleeping arrangements, medical conditions affecting sleep.
- Potential confounder: part-time night employment, which plausibly raises screen time and lowers sleep quality independently, meaning it could produce a correlation with no direct link between the two.
Note that "the effect of screen time on sleep quality" would overreach here. Nobody manipulated anyone's screen time, so the honest frame is relationship, and the IV and DV labels are conventional.
Worked Example 2: An Intervention Title
Illustrative.
Illustrative title: "Effect of Peer Tutoring on the Achievement of Grade 8 Students in Rational Algebraic Expressions."
- Independent variable: exposure to peer tutoring, a two-category grouping, tutored or not, or before and after in a single-group design.
- Dependent variable: achievement score on a teacher-made test.
- Control variables: the same teacher, the same topic coverage, the same test.
- Extraneous variables: prior mathematics grade, outside tutoring, absences during the intervention.
Here "effect of" is legitimate, because something was actually done. But if intact class sections were used rather than randomly assigned students, the design is quasi-experimental and the conclusion is worded more cautiously. If the topic itself is what you are teaching, our worked guide to rational algebraic expressions sets out the Grade 8 competencies the test would need to cover.
Worked Example 3: The Title That Breaks a Statistical Test
Illustrative.
Illustrative title: "Relationship Between Monthly Family Income and Senior High School Strand Chosen Among Grade 10 Completers."
- Independent variable: monthly family income, a ratio-level quantity in pesos.
- Dependent variable: strand chosen, a nominal category.
This is the case worth studying closely. Both variables are perfectly legitimate, the question is sensible, and yet the obvious test is unavailable. A Pearson correlation requires two continuous variables, and coding STEM as 1, ABM as 2 and HUMSS as 3 does not make strand continuous; the numbers carry no order or distance. The correct approach is to group income into ranges and run a chi-square test of independence, or to use a technique built for categorical outcomes.
Students discover this in Chapter 4, after collection, which is the worst possible time. Deciding the level of measurement for every variable before you print the questionnaire is a fifteen-minute task that prevents it.
Levels of Measurement Decide Your Test
| Level | What it does | Illustrative variable | Typical treatment |
|---|---|---|---|
| Nominal | Names categories, no order | Sex, strand, school type | Frequency, percentage, chi-square |
| Ordinal | Ordered categories, unequal gaps | Likert response, class rank | Median, mode, weighted mean by convention, Spearman rho |
| Interval | Ordered, equal gaps, no true zero | Standard score, temperature in Celsius | Mean, standard deviation, t-test, ANOVA, Pearson r |
| Ratio | Ordered, equal gaps, true zero | Hours of screen time, test score, income | All of the above |
One honest note about the ordinal row. Strictly, a Likert item is ordinal and its mean is not fully justified. Philippine research manuals almost universally accept weighted means of Likert items anyway, computed across multiple items and interpreted against a stated range table. Follow your manual, but know that this is a convention rather than a mathematical entitlement, and expect a statistics-minded panel member to raise it.
Operational Definitions: The Step That Decides Feasibility
An operational definition says how a variable is measured in your study. It is the bridge between an abstract concept and a column in your data file.
Illustrative operational definitions from the study above:
- Daily smartphone screen time: the weekly daily average in hours and minutes reported by the respondent's built-in screen time feature, converted to decimal hours.
- Self-reported sleep quality: the mean of five 4-point Likert items covering sleep onset, night waking, morning restedness, daytime alertness and daytime napping, with the napping item reverse-scored, interpreted against the range table stated in Chapter 3.
Read those and notice what they make possible: someone else could replicate the measurement exactly, the level of measurement is obvious, and the statistical test follows without argument. A definition that leaves any of those unclear is not finished.
Each operational definition should also point at specific items in your instrument. Our guide to building a research questionnaire covers that mapping in detail.
Common Mistakes That Cost Marks
- Swapping IV and DV in the title. "The effect of academic performance on study habits" reverses the plausible direction. Read your title in the "effect of A on B" frame and check that the arrow points the way you mean.
- Writing conceptual definitions and calling them operational. A dictionary sentence is not a measurement plan.
- Listing a moderator or mediator in the framework and never testing it. If it appears in the diagram, a sub-problem must address it.
- Treating a nominal variable as continuous. The single most expensive variable error in student research.
- Naming so many variables that none is measured well. Instruments have a practical length, and respondents have patience limits.
- Declaring control variables you did not actually control. If respondents came from three schools, school is not a control variable.
- Using IV and DV language in a qualitative study. Qualitative designs work with concepts and themes, not measured variables.
A Short Checklist
- Write your title, then rewrite it as "effect of A on B" or "relationship between A and B."
- Name the IV, the DV, the controls, and the extraneous variables you will declare.
- Write an operational definition for every variable, naming the specific items that measure it.
- Assign a level of measurement to each variable.
- Check that the statistical test you intend to run is legal at those levels. See analysing data in a student research paper.
- Confirm every variable appears in at least one sub-problem, and every sub-problem uses at least one variable.
Ready to find the right school? Browse schools or compare programs and reviews across hundreds of Philippine schools on SchoolFinderPH.
Disclaimer: terminology for intervening, mediating and moderating variables varies between Philippine research manuals. Use your own department's definitions where they differ from this guide.
Sources
- DepEd — K to 12 Senior High School Applied Track, Practical Research 2 Curriculum Guide — competency CS_RS12-Ia-c-3, "differentiates kinds of variables and their uses"
- DepEd — K to 12 Senior High School Applied Track, Practical Research 1 Curriculum Guide
- SchoolFinderPH — Statement of the Problem Guide
- SchoolFinderPH — Quantitative Research Designs
- SchoolFinderPH — Writing a Research Hypothesis
