7.1 The Nature, Role and Uses of Statistics in Social Work
Key Takeaways
- Topic A of the Social Welfare Policies TOS is Social Work Statistics and carries 10 of that paper's 100 items — a block most reviewers omit entirely.
- Descriptive statistics summarise data actually collected; inferential statistics generalise from a sample to a population.
- The four levels of measurement — nominal, ordinal, interval and ratio — determine which statistic is permissible.
- Social work uses statistics for needs assessment, targeting, programme monitoring, evaluation and policy advocacy.
- Calculators are prohibited inside the SWLE examination room, so items test reasoning about statistics rather than arithmetic computation.
7.1 The Nature, Role and Uses of Statistics in Social Work
Blueprint anchor. In the Enhanced Table of Specifications, the Social Welfare Policies, Programs and Services paper opens with A. SOCIAL WORK STATISTICS — 10%, followed by B. SOCIAL WORK RESEARCH 1 — 10% and C. SOCIAL WORK RESEARCH 2 — 10%. Thirty of the paper's 100 items are statistics and research. Candidates who prepare only legislation and programmes forfeit almost a third of this subject.
[!IMPORTANT] Examination reality. All kinds of calculators are prohibited inside the examination premises under the official PRC programme. Items therefore test conceptual command — which statistic is appropriate, what a result means, what a design permits — rather than long computation. Study accordingly.
1. What Statistics Is and Why Social Work Needs It
Statistics is the science of collecting, organising, presenting, analysing and interpreting numerical data to support decisions under uncertainty.
Social statistics applies that science to social phenomena — population, households, income, education, health, crime, service use — and is the evidentiary backbone of social welfare planning.
The Six Practice Uses
| Use | What statistics answers | Concrete Philippine example |
|---|---|---|
| Needs assessment | How many, where, how badly off? | Barangay profiling of out-of-school youth by sitio and sex |
| Targeting | Who should receive assistance? | Proxy means test scores used to rank household eligibility |
| Programme monitoring | Is delivery on track? | Compliance rates for schooling and health conditionalities |
| Evaluation | Did the programme change anything? | Comparison of outcomes before and after, or against a comparison group |
| Policy advocacy | What claim can be evidenced? | Income-loss estimates presented to a Local Development Council |
| Agency management | Where do resources go and with what result? | Caseload per worker, unit cost per service, waiting time |
2. Descriptive Versus Inferential Statistics
| Descriptive | Inferential | |
|---|---|---|
| Purpose | Summarise the data you actually have | Draw conclusions about a larger population from a sample |
| Question answered | "What does this set look like?" | "Can this finding be generalised, and how confidently?" |
| Typical tools | Frequency, percentage, mean, median, mode, range, standard deviation, tables, graphs | Estimation, confidence intervals, hypothesis tests, chi-square, t-test, ANOVA, correlation |
| Requires sampling theory? | No | Yes |
| Social work example | "Of 120 clients served this quarter, 68 percent were women." | "Clients in the group-work arm improved significantly more than those in the individual arm." |
3. Key Vocabulary
- Population (universe). Every unit about which a conclusion is wanted — for example, all senior citizens in a municipality.
- Sample. The subset actually studied.
- Parameter. A numerical characteristic of a population.
- Statistic. A numerical characteristic of a sample, used to estimate the parameter.
- Variable. A characteristic that varies across units — age, income, sex, service satisfaction.
- Independent variable. The presumed cause or grouping factor.
- Dependent variable. The presumed effect or outcome.
- Discrete variable. Counts only — number of children.
- Continuous variable. Any value within a range — income, hours of service.
4. Levels of Measurement
The level of measurement determines what may legitimately be computed. This is the most heavily tested statistical concept in the paper because it requires no arithmetic.
| Level | Property | Social work example | Permissible central tendency | Permissible operations |
|---|---|---|---|---|
| Nominal | Categories with no order | Sex, civil status, barangay, type of assistance | Mode only | Counting, percentage, mode, chi-square |
| Ordinal | Ordered categories, unequal intervals | Educational attainment, satisfaction rating, severity of risk | Median and mode | Ranking, percentile, median, Spearman rank correlation |
| Interval | Equal intervals, arbitrary zero | Standard scores, attitude scale scores | Mean, median, mode | Addition and subtraction; mean, standard deviation, Pearson correlation |
| Ratio | Equal intervals, true zero | Age, income, number of sessions, hours of field practice | Mean, median, mode | All arithmetic including ratios; "twice as much" is meaningful |
[!IMPORTANT] Classic distractor. Computing an arithmetic mean of nominal or ordinal data is a measurement error. "The average civil status is 2.3" is meaningless. If a stem reports a mean of a ranked satisfaction scale and asks what is wrong, the answer concerns level of measurement, not sample size.
5. The Social Worker as Producer, Not Only Consumer, of Statistics
Registered social workers generate the administrative data on which municipal and national planning rests: intake forms, case records, service outputs, referral logs, and profiling surveys. Three professional obligations follow:
- Accuracy at source. A poorly completed intake form propagates into a wrong municipal profile and a wrong budget allocation.
- Sex-disaggregation by default. The Enhanced TOS explicitly requires sampling and data collection using sex-disaggregated data, and gender-responsive planning is a legal requirement in Philippine programme design.
- Interpretation with context. Numbers do not speak; the worker supplies the social explanation. A rising number of reported abuse cases may indicate rising incidence, rising reporting confidence, or both, and the professional report must distinguish them.
6. Worked Practice Application
A municipal social welfare office reports the following for its crisis assistance programme in one quarter:
| Indicator | Value |
|---|---|
| Households assisted | 340 |
| Female-headed households assisted | 204 |
| Modal reason for assistance | Medical expense |
| Median assistance amount | ₱3,000 |
| Mean assistance amount | ₱4,850 |
| Highest single assistance | ₱50,000 |
Reading the table professionally:
- 204 of 340 is 60 percent female-headed — a descriptive percentage, appropriate for nominal data.
- "Medical expense" is the mode because reason for assistance is nominal; no mean is possible for it.
- The mean exceeds the median by a wide margin, which signals a right-skewed distribution pulled by a few very large grants. The median is the better summary of the typical household's experience.
- Reporting only the mean would overstate what an ordinary applicant receives — a misrepresentation with budget consequences.
- The finding that 60 percent of recipients are female-headed households belongs in the gender and development report and supports a case for targeted livelihood programming, but it does not by itself establish that women are more likely to need assistance, since the base rate of female-headed households in the municipality is not given.
That final caution — comparing a proportion of service users to a population base rate — is the single most useful statistical habit a practising social worker can carry into a case conference.
A report states that "the average educational attainment of clients is 2.7" where 1 = elementary, 2 = high school, 3 = vocational and 4 = college. What is the principal statistical defect?
A crisis assistance programme reports a median grant of ₱3,000 and a mean grant of ₱4,850. What does this pattern most likely indicate?
Which statement correctly distinguishes descriptive from inferential statistics?