Quantitative Research Methods for Social Scientists

This guide covers the basics of quantitative research methods for social scientists, including the four levels of measurements, measures of central tendency and dispersion, descriptive and inferential statistics, and interpretation of results obtained through SPSS output. It also explains important terms such as hypothesis and null hypothesis, independent and dependent variables, and extraneous variables. The guide concludes with references to further reading on the topic.

QUANTITATIVE
RESEARCH METHODS
FOR SOCIAL SCIENTISTS

TABLE OF CONTENTS
1. Four levels of measurements...................................................................................................3
2. Measure of central tendency and measure of dispersion........................................................3
3. Difference between descriptive and inferential statistics........................................................3
4. Explanation of terms...............................................................................................................4
5. Interpretation of results obtained through SPSS output..........................................................4
REFERENCES................................................................................................................................7

1. Four levels of measurements
1. Nominal: Categorization of data can be done by labelling them in groups that are
mutually exclusive. However, there is no order among categories. For example, Gender
and marital status (Hackett, 2018).
2. Ordinal: Along with categorization, ranking can be done of the data in order. However,
it cannot be said that how much intervals are there among ranks. For example, Top 5
medallists of Olympics and language ability (Fluent, intermediate or beginner).
3. Interval: Here, along with categorization both rankings and inference of equal intervals
among neighbouring data points could be done. However, there is no existence of true
zero point (Zyphur and Pierides, 2020). For example, Different within two adjacent
temperature is one degree, but the definition of zero degree is different and depends on
the scale which doesn't mean absence of temperature absolutely.
4. Ratio: It is different from interval level in terms of existence of true zero point. For
example, in Kelvin temperature scale - zero signifies absolute absence of thermal energy.
2. Measure of central tendency and measure of dispersion
Measures of Central tendency: There are mainly three measures through which central
tendency can be determined of the given data set that is, mean, mode & median. These measures
provide with single value which is meant for describing the entire data set through identifying its
central position within the given data set (Liamputtong, 2019).
Measures of dispersion: There are five commonly used measures to determine the level of
dispersion among the data set that is, range, variance, standard deviation, mean deviation and
interquartile range. The value describe how much the data points within the data set has been
spread from each other.
3. Difference between descriptive and inferential statistics
Descriptive Statistics Inferential Statistics
It aims to describe the characteristics of the
target population (Ewing and Park, 2020).
It aims to make inferences on the basis of
sample and then generalise them for the entire
population.

It helps in organization, analysis and
presentation of data set in meaningful way.
Here comparison and testing is done to predict
future outcomes.
4. Explanation of terms
a. The hypothesis and null hypothesis: Hypothesis is the proposed explanation that is being
made for a phenomenon. It can be referred to as either an idea or assumption which is then tested
to determine whether the proposed argument is true or not.
A null hypothesis is statistical theory which proposes that there is no statistical
significance lies within the given set of observed data (Richards and et.al., 2019).
b. Independent & dependent variable: Independent variable can be regarded as the cause
whose value is not dependent on any other variables forming part of the study. However,
dependent variable can be regarded as the effect whose value depends on changes taking place in
independent variable.
c. Extraneous variables: These are the variables that are not taken into account during
investigation which are meant to have potential impact on the outcomes of the research
conducted, If it is not controlled then there would be inaccurate conclusions regarding the
association among dependent & independent variables (Tominc and et.al., 2018).
5. Interpretation of results obtained through SPSS output
a. Frequency table
The output chart headed by the name of variable under consideration that is, “Noisy neighbours /
loud parties” indicates frequency distribution of it. There are four options available to the
respondents to choose while collecting data from them that is; Very common, fairly common, not
very common and not at all common and the same are coded as 1,2,3,4, so that interpretation of
other statistical results could be done on the basis of these numerical values assigned to each
options.
The analysis consists of total observations equals to 3143 and the highest observation is
of 'not at all common' with 1635, then 'not very common' as 1105, 'fairly common' as 255 and
'very common' as 148 (Hackett, 2018.).
The percent column indicates proportion of each option on the basis of whole data which
includes missing observations while the valid percent indicates proportion of each option on the

basis of total of non-missing observations only. Cumulative percent are helpful for determining
cut-offs for quartiles.
b. Cross-tabulation
The raw variable indicates CommonnessOfHavingNoisyNeighbours while the column variable
indicates AgeGrouped. Accordingly, column percentages are meant for what percentage of
people belonging from a particular age group would commonly be noisy neighbours. Like,
variable AgeGrouped will be helpful in determining the denominator while computing
percentage (Bloomfield and Fisher, 2019).
 The proportion of sample belonging from young age group who are very common to be a
noisy neighbours is 6.6%.
 The proportion of sample who are young and fairly common to be loud parties is 11.6%.
 The proportion of people being young and not very common to be noisy neighbours is
40.7%.
 The percentage of sample being young and not at all found to be loud parties is 41.2%
 The proportion of adults who are very common be noisy neighbours is 5.5%.
 The percentage of adults who are fairly common to be noisy neighbours is 8.2%.
 The proportion of adults who are not very common to be noisy is 37.1%.
 The proportion of adults who are not at all found to be loud parties is 49.3%.
 The proportion of elderly people who are very common to be noisy is 2.2%.
 The percentage of elderly people who are fairly common to be noisy neighbours is 5.9%.
 the proportion of people belonging from elderly age group and are not very common to
be noisy is 28.3%.
 The percentage of people belonging from elderly age group and are not at all found to be
loud parties is 63.6%.
c. Chi-square result
The value of chi-square statistics appears in the value column corresponding to Pearson Chi-
square that is, 91.182 in the given results. Here p value is less than 0.05 and thus, null hypothesis
would be rejected while alternative hypothesis will be accepted which states that there is
statistical significance within people from different age group and their being noisy neighbours
(Zyphur and Pierides, 2020).
d. Phi / Cramer's

These are the tests performed to determine the strength of association between the variables.
Therefore, phi = 0.171 and Cramer's V = 0.121 indicates that the strength of association between
age group and noisy neighbours is quite strong.

REFERENCES
Zyphur, M. J. and Pierides, D. C., 2020. Statistics and probability have always been value-laden:
An historical ontology of quantitative research methods. Journal of Business
Ethics, 167(1), pp.1-18.
Bloomfield, J. and Fisher, M. J., 2019. Quantitative research design. Journal of the Australasian
Rehabilitation Nurses Association, 22(2), pp.27-30.
Hackett, P. ed., 2018. Quantitative research methods in consumer psychology: Contemporary
and data driven approaches. Taylor & Francis.
Tominc, P., and et.al., 2018. Students’ behavioral intentions regarding the future use of
quantitative research methods. Naše gospodarstvo/Our economy, 64(2), pp.25-33.
Richards, K. A. R., and et.al., 2019. Studying recruitment and retention in PETE: Qualitative and
quantitative research methods. Journal of Teaching in Physical Education, 38(1), pp.22-
36.
Ewing, R. and Park, K. eds., 2020. Basic Quantitative Research Methods for Urban Planners.
Routledge.
Liamputtong, P. ed., 2019. Handbook of research methods in health social sciences. Singapore::
Springer.