Research and Statistics Flashcards

Question

Relationship Education Evidence Base for MFT

Answer 1

Strong evidence of **improvement with non-distressed couples**.

Answer 2

Couples therapy indicated when **marital problems** are part of presenting issues.

Answer 3

Strong evidence for **family therapy**, especially for **diabetes** and **asthma**.

Answer 4

**RELIABILITY** = consistency. If you measure something repeatedly, do you get the same result? A reliable scale gives you the same weight each time. **VALIDITY** = accuracy. Does the test actually measure what it claims to measure? A scale that's consistently 10 pounds off is reliable but not valid. ## Footnote Memory trick: Reliable = Repeatable. Valid = on the V-mark (right on target).

Answer 5

Research shows that outcome variance can be attributed to: * 40% client variables (e.g. motivation, resources, severity, etc.) * 30% quality of the therapeutic relationships * 15% hope/placebo * 15% therapeutic model

Answer 6

Descriptive statistics **SUMMARIZE** and **DESCRIBE** data. They paint a picture of what you're looking at without making predictions. Common types: * Mean (average) * Median (middle value) * Mode (most common) * Range (highest to lowest) * Standard deviation (how spread out the data is) ## Footnote Example: If you're tracking session attendance rates, descriptive stats tell you the average number of sessions clients attend, the range, etc.

Answer 7

Using your sample data to make **predictions** or draw **conclusions** about a larger population (what MIGHT BE). ## Footnote Example: "Based on our sample, we can infer that the average attendance for all clients in this population is likely between 7.5 and 9.5 sessions, with 95% confidence"

Answer 8

It compares the means of **TWO** groups to see if they're significantly different. Common uses: * Independent t-test: Compare two different groups (treatment vs. control) * Paired t-test: Compare same group at two times (pre-test vs. post-test) ## Footnote Example: Do couples who complete therapy (Group A) have significantly different relationship satisfaction scores than couples on a waitlist (Group B)? Remember: t-test is for comparing TWO groups. For 3+ groups, use ANOVA.

Answer 9

**ANOVA** (Analysis of Variance): Compares means across 3+ groups to see if there are significant differences. **ANCOVA** (Analysis of Covariance): Same as ANOVA BUT controls for a confounding variable (covariate). ## Footnote Think: ANCOVA = ANOVA + Control variable ANOVA Example: Does therapy outcome differ between CBT, DBT, and family therapy groups? ANCOVA Example: Does therapy outcome differ between the three groups WHILE controlling for initial symptom severity?

Answer 10

A 95% confidence level means if you repeated your study 100 times, about 95 of those times your results would fall **within the same range**. ## Footnote What it tells you: How confident you can be that your findings aren't just random chance. Common in research: 95% is standard (sometimes 99% for more certainty). Translation: 'We're 95% sure this result is real, not a fluke.'

Answer 11

It measures the **strength** and **direction** of a relationship between two variables. Range: -1.0 to +1.0 * +1.0 = perfect positive correlation (both go up together) * 0 = no correlation * -1.0 = perfect negative correlation (one goes up, other goes down) ## Footnote Example: Correlation between number of therapy sessions and symptom reduction might be r = 0.65 (strong positive relationship). Remember: Correlation ≠ causation! Just because things relate doesn't mean one causes the other.

Answer 12

An evidence-based treatment has been rigorously **tested** through research and shown to be **effective** for specific conditions. Three components (Sackett's definition): 1. Best research evidence (controlled studies) 2. Clinical expertise (your professional judgment) 3. Client values/preferences (what matters to them) ## Footnote Important: EBT doesn't mean ignoring your clinical judgment or the client's context—it means integrating all three. Examples: EFT for couples, MST for adolescents, MDFT for substance abuse

Answer 13

**LINEAR CAUSALITY**: A causes B (straight line, one-way, simple cause-and-effect). Example: "Mom's depression causes the child's behavior problems." Assumes: Single direction, one factor responsible. **CORRELATION**: Two variables are related or associated, but we can't assume one causes the other. Example: "There's a correlation between mom's depression and child's behavior problems" Means: They occur together, but we don't know: * Does A cause B? * Does B cause A? * Does a third factor (C) cause both? * Is it coincidence? In family therapy: We often see correlations between family members' symptoms and behaviors. Understanding that correlation ≠ causation helps us avoid oversimplifying complex family dynamics.

Answer 14

**MANOVA** (Multivariate Analysis of Variance): Like ANOVA, but compares 3+ groups on MULTIPLE dependent variables simultaneously. **MANCOVA** (Multivariate Analysis of Covariance): Same as MANOVA BUT controls for one or more confounding variables (covariates). ## Footnote Key difference from ANOVA/ANCOVA: * ANOVA/ANCOVA = ONE outcome variable * MANOVA/MANCOVA = MULTIPLE outcome variables tested together Why use it: More efficient than running separate ANOVAs, and it accounts for correlations between outcome variables. MANOVA Example: Does therapy type (CBT, DBT, Family Therapy) affect depression scores, anxiety scores, AND relationship satisfaction scores all at once? MANCOVA Example: Does therapy type affect those three outcomes WHILE controlling for baseline symptom severity and number of sessions attended?

Answer 15

It examines the **relationship** between **one or more independent variables** (predictors) and a **dependent variable** (outcome). It helps you understand: 1. IF there's a relationship 2. HOW STRONG the relationship is 3. The DIRECTION of the relationship ## Footnote Example: Can we predict therapy outcomes (dependent variable) based on number of sessions, client age, and initial symptom severity (independent variables)? Key difference from correlation: Regression allows prediction and examines multiple predictors simultaneously, while correlation just shows if two variables are related.

Research and Statistics Flashcards

Learn key research methods, data analysis concepts, and statistical tools used to evaluate outcomes and evidence in marriage and family therapy. (39 cards)