Connor Whiteley GMBPsS
Psychological Statistics:
Welcome to the Psychology Statistics section of the website!
​
Personally, I quite like statistics but it doesn't mean I find it easy!
​
So, here are a list of psychology statistic questions and answers that I struggled with at first from both my psychology undergraduate degree and Clinical Psychology MSc. Hopefully, these statistical questions will help you on your psychology journey.
​
Enjoy!
Psychology Statistic Questions I Struggled With During my Undergraduate Degree
What is a Type 1 Error?
Please scroll down to the next section on this statistics webpage to see the questions I struggled with during my Clinical Psychology MSc.
A type 1 error is the rejection of a true null hypothesis.
What is a (alpha) Determined By In Statistics?
a is determined by the researcher.
When are the Differences of The Means most relevant in Psychology Statistics?
The difference of the means is most relevant in an independent groups t-test analysis.
What is not an assumption of a t-test?
Groups having different means is NOT an assumption of an independent t-test.
What does a null hypothesis mean for an ANOVA?
A null hypothesis for an ANOVA is that none of the means are significantly different.
When Do You Run A One-Way ANOVA in Psychology?
If you have a study with one independent variable and 3 levels, you need to run a single-factor ANOVA.
What Tests Can A Bonferroni Test Be Used On?
A Bonferroni test can be run on pairwise pre-planned tests and pre-planned tests only.
What Does Orthogonal Mean In Statistics?
Orthogonal means having variables that can be treated independently.
How Do You Make Contrasts Orthogonal In Psychology Statistics?
For making contrasts orthogonal, you don’t have to have equated contrasts nor products of all the alternative conceptional contrasts equal zero.
What is an F Statistic?
An F statistic is the value you get after running an ANOVA or regression analysis. This tells you if the means between the two population are significantly different.
When Is a Two Way ANOVA Used?
A two-way ANOVA is used to analyse data from a study with two independent variables and as many levels as necessary.
What is a Pairwise Analysis?
A pairwise analysis is a comparison between two levels of one IV level and a comparison between two conditions.
How Can You Increase Statistical Power?
You can increase the statistical power by increasing a, increasing the sample size and increasing the difference between the means.
Does Increasing the Variability of The Population Increase Statistical Power?
Increasing the variability of the population does not increase the statistical power of a psychology study.
What Does a (alpha) Test in Statistics?
a (alpha) is the significance level in hypothesis test or acceptable probability of a type 1 error.
What is Standard Deviation?
Standard deviation is the square root of variance.
When is a One -Way ANOVA Used?
A one way ANOVA is used on data with 1 independent variable or factor with as many levels as necessary.
What Type of Test is The Tukey’s ‘Honesty’ Significant Difference test?
The Tukey’s ‘Honesty’ significant difference test is a post-hoc test.
What is a Grand Mean?
A Grand Mean is the average of the means of other cells or subgroups within the dataset.
How Do You Compute a New Variable In SPSS?
In SPSS, to compute a new carriable you need to go to the Transform on the top bar. Then you need to select ‘Compute Variable’
What is a t-Statistic?
The t-statistic is used in t-tests to determine where to support or reject the null hypothesis.
How Do You Run Main Effects and Interaction Effects In SPSS?
To run main effects and interaction effects in SPSS, you need to run an ANOVA by going to SPSS analyse, General Linear Model and Univariate. Then you enter your IVs in the fixed factor box and your DV in the DV box.
Psychology Statistic Questions I Struggled With During my Masters Degree
What Is Nominal Data?
Nominal data is the most basic form of data and this is for labelling variables. For example, types of mental health conditions.
​
What Is Ordinal Data?
Ordinal data is about order and ranking variables.
What Is Interval Data?
Interval data are numbers without a true 0. Like IQ.
What Descriptive Statistic Is Most Affected By Outliers In Data?
The mean is most affected by outliers.
What Is Positively Skewed Distribution Or Data?
When something is positively skewed, the long tail is longer on the right because it is running up the positive numbers.
What Is Negatively Skewed Distribution Or Data?
When something is negatively skewed, the long tail is longer on the left because it is running down the negative numbers.
What Is A Reference Category In Regression?
In a General Linear Model, a reference category is the category all other categories are compared against. This is set up by the user, or it is the category with the highest value.
What Is R-Squared?
In regression, R-squared is a measure of how well the linear regression fits the data. This is given as a percentage from 0% to 100%.
How To Know How Many To Subtract N By In Degrees Of Freedom?
In degrees of freedom, to know whether it is N-1 or N-X (X meaning whatever number). You need to look at how many parameters you have. A parameter is a number that describes a whole number. Therefore, if you’re looking at Mean and Standard Deviation then you have two parameters.
What is Central Limit Theorem?
The basic point of Central Limit Theorem is that given enough sample participants even the most skewed data can give you a normal distribution.
What Is A Type 1 Error In Statistics?
A false positive is a type 1 error.
What Is A Type 2 Error In Statistics?
A false negative is a type 2 error.
What Is Alpha In Statistics?
Alpha is the significance or probability of rejecting the null hypothesis when it is true. like a significance of 0.05.
What Is Beta In Statistics?
Beta is the probability of us committing a type 2 error (a false negative).
What Are Some Criticisms Of Significance Testing?
4 criticisms of significance testing includes:
-
Null hypothesises are almost always false in reality because there are always “some” differences that will show up given enough data. Like drug A and B will never be 100% identical so some differences will be found.
-
Small p-value doesn’t mean small chance of being wrong because in 1 review 74% of studies with a small p-value were wrong. (Granted I got this criticism from a source in the early 2000s so this probably has a temporal validity issue).
-
Statistical error is only 1 small part of real error.
-
Statistical significance is not the same as scientific significance. Science deals with questions of effect size, not whether an effect exists.
​
What Is The Error Testing Approach To Significance Testing Called?
The error testing approach to significance testing is called Neyman-Pearson.
​
What Is P-Value Probability?
The p-value probability is you could get the result like the one observed if the null hypothesis was true.
What Can Influence Effect Size?
Mean difference, methodology and variability can all influence effect size but sample size does not.
What is Multiple R?
Multiple R is the correlation between X and Y.
What is Multiple R-Squared?
Multiple R-squared is the proportion of variance in Y that is accounted by the regression equation.
What Are b And beta Coefficients?
The b and beta coefficients are used in regression to measure the relationship between the dependent and one or more independent variables. b is used to represent one unit change in an independent variable. Whereas beta coefficients represent 1 change in standard deviation for the independent variables.
What Is Least Square Criterion?
The Least Square Criterion is a test used to test the accuracy of a straight line in depicting the data it was made from.
What Is Statistical Power?
Statistical power is the probability that a test will detect an effect or difference in data when it exists.
What Is A Power Analysis Made Up Of?
A power analysis is made up of the effect size, the power, the significance level and the sample size and one of its components.
What Can Power Analysis Be Used For?
You can use a power analysis for:
-
Comparing different statistical tests and choosing the test with the highest power to detect an effect of interest.
-
Calculating the optimal sample size for a given sample effect size and significance level, and vice versa.
-
Calculating the minimum sample size needed to achieve a certain level of statistical power and significance level.
What Is The Recommended Effect Size In Psychology Statistics?
80% is the generally recommended effect size in psychology.
What Do Sensitivity Tests Measure?
Sensitivity tests measure how well a test can identify a presence or absence of a particular outcome or condition.
What Do One-Way ANOVAs Analyse?
A one-way ANOVA analyse the difference between two or more groups.
What Data Does A Chi-Square Test Use?
A chi-square test uses categorical data.
What are F-Ratios?
The f-ratio measures the variance between group means and compares it to the variance within groups. Higher f-ratio values mean there are greater differences whereas lower f-ratio values mean there are lower or no differences.
What are F-Tests?
An F-test compares the variances of 2 or more groups of data based on F-ratio. F-tests can be used for testing the good fit of a regression model, testing the difference between 2 or more means in an ANVOA, or testing the equality of variances.
​
What Is A Sum Of Squares Error?
A sum of squares error is a calculation that measures the difference between predicted and actual values. It’s used to evaluate the accuracy of a regression model where a lower SSE shows the model is a better fit for the data.
What Are Type 1 Sum Of Squares?
Type 1 sum of squares is where all the variances of the main effects gives to the first main effect entered.
What Are Type 3 Sum Of Squares?
Type 3 sum of squares is where all the shared variance amongst the main effects and interactions is discarded, so all the main effects and interaction control for each day.
What Are Type 2 Sum Of Squares?
Type 2 sum of squares gets rid of the shared variance of the main effects, but not the shared variance between the main effects and interactions.
What Are Main Effects In Statistics?
Main effect is the effect of an independent variable on a dependent variable.
When Do You Use A Factorial ANOVA?
You use a factorial ANOVA when you want to look at 2 or more nominal independent variables on an ordinal dependent variable.
What Do Main Effects Look At?
Main effects look at differences amongst marginal effects.
What Do Interactions Look At?
Interactions look at the differences amongst cells that aren’t accounted for by the main effects.
​
What Do Simple Main Effects Look At?
Simple main effects look at one factor within one level of another factor.
What Is The Name Of Qualitative Research?
Qualitative research aims to explore the meaning participants give to their experiences.
What Is The P-Value In Statistics?
The p-value is the probability you could get a result like the one observed if the null hypothesis was true.
Degrees of Freedom Information
When it comes to degrees of freedom reporting questions, look at how many things are being measured. Like weight and height are two things so it is n-2.
How Is Variance Assigned In Hierarchical Analysis?
In hierarchical analysis, variance shared between multiple independent variables and the dependent variable is assigned to the independent variables that come first causally.
How Many Variables Take Part In A Partial Correlation?
3 variables take part in a partial correlation.
What Is Error Variance?
Error variance is the portion of variance in a datasheet due to extraneous variables and measurement error.