Behavioural Statistics-II

Keziah Mary Sam

Module 1:Introduction to hypothesis testing
Research Hypothesis, Rationale And Purpose Of Hypothesis Testing,
Estimation In Inferential Statistics, The Probability Value, Null And
Alternate Hypothesis. Critical Values, P-Values And Significance
Level, One And Two Tailed Tests, Steps In Hypothesis Testing, Errors In
Hypothesis Testing: Type I & II Errors, Effect Size. z-Test And t-Test.
Statistics
Statistics is the science of classifying, organising, and analysing data (King, Rosopa,
Minium, 2018)

The term statistics refers to a set of mathematical procedures for organising,

summarising, and interpreting information (Gravetter, Wallnau, Forzano, 2021)
Levels of Measurement
Measurements are “rules for assigning numbers to objects in such a way as to
represent quantities of attributes” (Nunnally,1978).
Measurement “consists of rules for assigning symbols to objects to (1) represent
quantities of attributes numerically (scaling) or (2) define whether the objects fall in
the same or different categories with respect to a given attribute (classification)”
(Nunnally & Bernstein, 1994)

1. Nominal
2. Ordinal
3. Interval
4. Ratio
Research Hypothesis
A research hypothesis is a statement of an expected or predicted relationship between
two or more variables.
It's what the experimenter believes will happen in their research study

Directional hypothesis
Directional hypothesis are those where one can predict the direction (effect of one
variable on the other as ‘positive’ or ‘negative’)

Non Directional hypothesis

A non-directional hypothesis is a two-tailed hypothesis that does not predict the direction
of the difference or relationship
Statistical hypothesis
A statement about the nature of a population. It is often stated in terms of a population parameter

Null hypothesis
A null hypothesis is a hypothesis that says there is no statistical significance between the two
variables.
It is usually the hypothesis a researcher or experimenter will try to disprove or discredit.
There is no difference between girls and boys in IQ level

Alternative hypothesis
It is a statement that contradicts the null hypothesis, suggesting that there is a significant
effect or difference.
Boys and girls are different in their empathy
Hypothesis Testing
Hypothesis testing is a statistical method that uses sample data to evaluate a hypothesis
about a population.
Hypothesis testing is a systematic way to test claims or ideas about a group or population.
Hypothesis testing or significance testing is a method for testing a claim or hypothesis
about a parameter in a population, using data measured in a sample.
In this method, we test some hypotheses by determining the likelihood that a sample statistic
could have been selected, if the hypothesis regarding the population parameter were true.
Rationale and Purpose of Hypothesis Testing
Making Inferences from Samples: In many practical situations, it's impractical or impossible
to collect data from entire populations. Instead, researchers collect samples from populations
and use statistical methods to make inferences about population parameters based on these
samples.

Null Hypothesis vs. Alternative Hypothesis: The core of hypothesis testing lies in setting up
two competing ideas:
The null hypothesis (Ho): This is the default assumption, often stating that there's no
effect or difference between things being compared.
The alternative hypothesis (Ha): This is the opposite of the null hypothesis, what we're
actually interested in finding evidence for.
Rationale and Purpose of Hypothesis Testing
Statistical Significance: If the data shows an effect so unlikely to occur by random chance
(often measured by a p-value), we can then reject the null hypothesis. This strengthens the
case for the alternative hypothesis, providing evidence that there's likely a real effect or
difference at play.

Data-Driven Decisions: By following a structured hypothesis testing framework, researchers

can avoid basing conclusions on hunches or biases. The analysis provides a level of
confidence in the results, allowing researchers to make more informed decisions about their
field of study.

Assessment of Claims: Hypothesis testing provides a systematic framework for assessing

claims or hypotheses about population parameters. These claims could be about the mean,
proportion, variance, or other characteristics of a population.
Rationale and Purpose of Hypothesis Testing
Quantifying Uncertainty: Hypothesis testing allows researchers to quantify the uncertainty
associated with their conclusions. By calculating probabilities and establishing thresholds for
rejecting or failing to reject hypotheses, researchers can make informed decisions while
acknowledging the inherent uncertainty in statistical inference.

Testing Research Questions: Researchers often have specific research questions or

hypotheses that they want to investigate. Hypothesis testing provides a formalized approach
to evaluate these hypotheses using empirical data.
Rationale and Purpose of Hypothesis Testing
Control of Type I and Type II Errors: Hypothesis testing helps control two types of errors:
Type I errors (false positives) and Type II errors (false negatives). By specifying a significance
level (α), researchers control the probability of making a Type I error, thereby ensuring the
reliability of their conclusions.

Scientific Rigor and Reproducibility: Hypothesis testing is a fundamental aspect of

scientific inquiry, promoting rigour and reproducibility in research. By following standardized
procedures for hypothesis testing, researchers ensure that their findings are transparent,
replicable, and subject to scrutiny by the scientific community.
Inferential Procedures
There are two types of inferential procedures :
1. Estimation
2. Hypothesis testing
Estimation In Inferential Statistics
In estimation, a sample is drawn and studied and inference is made about the population
characteristics based on what is discovered about the sample. There may be sampling
variations because of chance fluctuations, variations in sampling techniques, and other
sampling errors.
We, therefore, do not expect our estimate of the population characteristics to be exactly
correct.
We do, however, expect it to be close.
The real question in estimation is not whether our estimate is correct or not but how close is it
to the true value.
Characteristics of X as an estimate of (µ)
The sample mean (x̄ ) often is used to estimate a population mean (µ).
For example, the sample mean of 45.0 from the Academic Anxiety Test may be used to
estimate the mean Academic Anxiety of the population of college students. Using this sample
would lead to an estimate of 45.0 for the population mean. Thus, the sample mean is an
unbiased and consistent estimator of the population mean.

The goal of estimation is to estimate a value for an unknown population characteristic

(parameter) based on a sample statistic.
Both point estimators and interval estimators can be used for this task.
Unbiased Estimator:
An unbiased estimator is one in which, if we were to obtain an infinite number of random
samples of a certain size, the mean of the statistic would be equal to the parameter. The
sample mean, (x̄ ) is an unbiased estimate of (µ) because if we look at possible random
samples of size N from a population, the mean of the sample would be equal to µ.
Consistent Estimator:
A consistent estimator is one that as the sample size increases, the probability that the
estimate has a value close to the parameter also increases. Better estimates of a population
mean should be more probable from large samples.
Accuracy of Estimation:
The sample mean is an unbiased and consistent estimator of (µ). But we should not overlook
the fact that an estimate is just a rough or approximate calculation. It is unlikely in any
estimate that (x̄ ) will be exactly equal to (µ). Whether or not x̄ is a good estimate of (µ)
depends upon the representativeness of the sample, the sample size, and the variability of
scores in the population.
Point Estimation
We have indicated that x̄ obtained from a sample is an unbiased and consistent estimator of
the population mean (µ).
Thus, if a researcher obtains an Academic Anxiety Score from 100 students and wants to
estimate the value of (µ) for the population from which these scores were selected, the
researcher would use the value of x̄ as an estimate of (µ). If the obtained value of x̄ was 45.0,
this value would be used as an estimate of (µ). This form of estimate of population
parameters from sample statistics is called point estimation. Point estimation is estimating
the value of a parameter as a single point
for example, (µ) = 45.0 from the value of the statistic x̄ = 45.0
Interval Estimation
A point estimate of the population mean almost is assured of being in error, the estimate from
the sample will not equal to the exact value of the parameter.
To gain confidence about the accuracy of this estimate we may also construct an interval of
scores that is expected to include the value of the population mean. Such intervals are called
confidence intervals.
A confidence interval is a range of scores that is expected to contain the value of (µ).
The lower and upper scores that determine the interval are called confidence limits.
A level of confidence can be attached to this estimate so that the researcher can have a 95%
or 99% confidence level that encompasses the population mean.