LESSON 1.

PRESENTATION, ANALYSIS AND INTERPRETATION OF DATA

PRE-TEST:
1. Describe the instrument/s that you will use for your research study and detail your
justification.
2. Enumerate the statistical techniques you will use and the justification for using such.

Learning Activities:
The learner shall be able to:
1. Collects data using appropriate instruments.
2. Presents and interprets data in tabular and graphical forms.
3. Uses statistical techniques to analyze data.

Data collection can be highly formalized and controlled, or it can be as simple as

watching, and listening. Many researchers find it useful to collect information and data using
multiple techniques to provide a complete answer to a question. Generally, data and/or
obtaining and examining materials.
COLLECTING DATA USING APPROPRIATE INSTRUMENTS
Quantitative research is perhaps the simpler to define and identify. The data produced
are always numerical, and are analyzed using mathematical and statistical methods. If there
no numbers involved, then it’s not quantitative research.
Some phenomena obviously lend themselves to quantitative analysis because they are
already available as numbers. However, even phenomena that are not obviously numerical in
nature can be examined using quantitative methods.
The development of Likert scales and similar techniques mean that most phenomena
can be studied using quantitative techniques. This is particularly useful if you in an
environment where numbers are highly valued and numerical data is considered the gold
standard.
If you choose an existing instrument it is essential to do a quality control using
reliability, validity and relevance. Do not be afraid to contact the authors with questions about
supplementary information.
 If you do not find any appropriate measuring instruments you will have to develop
your own instruments. This is not easy. You cannot just make up questions for a
questionnaire. Your research question will be your starting point and your questions will have
to be based on strong theoretical arguments.
But how does a researcher determine the appropriate instrument or combination of
instruments that would answer his research problem? In choosing an instrument or
combination of instruments to do collect needed data you have to consider what is the most
appropriate instrument/s for investigating a particular research problem or question.
In reviewing what each of them is best used by taking into account their advantages
and disadvantages is part of the process of designing your own research study.
Survey Instruments
Among the different instruments of data gathering for research purposes, the survey
questionnaire is preferred by many researchers.
Advantages of survey questionnaires:
a. High representativeness – Survey provide a high level of general capability in
representing a large population. Due to the usual huge number of people who answers
the survey, the data being gathered possess a better description of the relative
characteristics of the general population involved in the study.
b. Low costs – When conducting surveys, you only need to pay for the production of
survey questionnaires. If you need a larger sample of the general population, you can
allot an incentive in cash or kind. On the other hand, other data gathering methods
such as focus groups and personal interviews require researchers to pay more.
c. Convenient data gathering – Surveys can be administered to the participants through
a variety of ways. The questionnaires can simply be sent via e-mail or fax, or can be
administered through the internet. Nowadays, the online survey method has been the
most popular way of gathering data from target participants.
d. Good statistical significance – Because of high representativeness brought about by
the survey method, it is easier to find statistically significant result than other data
gathering methods. Multiple variables can also be effectively analyzed using surveys.
e. Little or no observer subjectivity – Surveys are ideal for scientific research studies
because they provide all the participants with a standardized stimulus. With such high
reliability obtained, the researcher’s own biases are eliminated.
 f. Precise results – As questions in the survey should undergo careful scrutiny and
standardization, they provide uniform definitions to all the subjects who are to answer
the questionnaires. Thus, there is no greater precision in terms of measuring the data
gathered.
Disadvantages of survey questionnaires:
a. Inflexible design – The survey that was used by the researcher from the very
beginning, as well as the method of administering it, cannot be changed all throughout
the process of data gathering. Although this inflexibility can be viewed as a weakness
of the survey method, this can also be strength considering the fact that preciseness
and fairness can both be exercised in the study.
b. Not ideal for controversial issues – Questions that bear controversies may not be
precisely answered by the participants because of the probably difficulty of recalling
the information related to them.
c. Possible inappropriateness of questions – Questions in surveys are always
standardized before administering them to the subjects. The researcher is therefore
forced to create questions that are general enough to accommodate the general
population. However, these general questions may not be as appropriate for all the
participants as they should be.
INSTRUMENT DEVELOPMENT
No research study can achieve success without a well-designed and a well-thought
instrument. Unfortunately, instrument design has no theoretical base to guide the researcher
in developing a flawless instrument. All the researcher has to guide them is a lengthy list of
do’s and don’ts born out of the experience of other researchers past and present. Hence,
instrument design is more of an art than a science.
Survey Questionnaire
A survey questionnaire or instrument is a tool for consistently implementing a
scientific protocol for obtaining data from respondents. For most social and behavioral
surveys, the instrument involves a questionnaire that provides a script for presenting a
standard set of questions and response options. The survey instrument includes questions
that address specific study objectives and may also used to collect demographic information
for calculating survey weights. In some surveys, questionnaire responses are augmented by
other kinds of measurements derived from instruments, such as lab samples or physical
assessments.
A difficult task in creating a survey questionnaire is translating a researcher’s
questions into items that are sufficiently succinct and simple for the respondent to
understand and provide accurate answers.
In general, a survey question should:
a. Contain only one idea or question
b. Define the scope to consider, such as the time period or activities that are relevant
to the question
c. Be written with neutral language to avoid leading the respondent to a specific
answer
d. Use language that enables less educated persons to easily understand the
question
e. Contain response options that are simple, clear, consistent, and include the full
range of responses that might occur
f. For categorical responses, be mutually exclusive and exhaustive so that a
respondent can pick one and only one option.
g. For numeric responses, guide the respondent to provide the response in
consistent format and units
Questionnaires may be designed as:
1. Structured questionnaires – They are those in which some control or guidance is given
for the answer. This may be described as closed form because the questions are
basically short, requiring the respondent to provide a “yes” or “no” response, or
checking an item out of a list of given responses. Questions that require yes or no
answers are also termed as Dichotomous questions. It may, also be multiple choice
options from which the respondent selects the answer closer to their own opinion.
The respondent’s choices are limited to the set of options provided.
2. Unstructured questionnaire – This type which is also termed as open-ended or
unrestricted type of questionnaire calls for a free response in the respondent’s own
words. The respondent frames and supplies the answer to the question raised in the
questionnaire. It also constitutes questions which give the respondent an opportunity
 to express his or her opinions from a set of options. Spaces are often provided for
respondents to make their inputs.
In general, it is best for a survey questionnaire to be as short as possible. A long survey
questionnaire leads to a long interview and this is open to the dangers of boredom on the
part of the respondents (and poorly considered, hurried answers), interruption by third
parties and greater costs in terms of interviewing time and resources. In a rural situation an
interview should not last longer than 30-45 minutes.
Questions to be Avoided in a Survey Questionnaires:
There is something more important than knowing the questionnaire format and what
type of questions to be asked in the questionnaire. It is understanding, what questions need
to be avoided in a survey. Take care to avoid following type of questions when preparing a
questionnaire:
1. Embarrassing questions – Questions that ask respondents details about their personal
and private matters are embarrassing questions. Such types of questions are better to
be avoided as you risk losing trust of your respondents. Your respondents might also
feel uncomfortable to answer such questions and might refuse to answer your
questionnaire altogether.
2. Positive/negative connotation questions – Since most verbs, adjectives and nouns in
the English language have either a positive or negative connotation, questions are
bound to be taken as either positive or negative. While defining a question, strong
negative or positive overtones must be avoided. Ideal questions should have neutral
or subtle overtones.
3. Hypothetical questions – Hypothetical questions are based on speculation and
fantasy. An example of a hypothetical question would be “If you were the CEO of an
4. ABC organization what would be the changes that you would bring?” Questions such
as these, force respondents to give their ideas on a particular subject, and generally
the data collected through such questions are inconsistent and unclear. Hypothetical
questions should be avoided in questionnaires.
Piloting the questionnaires
Even after the researcher has proceeded along the lines suggested, the draft survey
questionnaire is a product evolved by one or two minds only. Until it has actually been used
in interviews and with respondents, it is impossible to say whether it is going to achieve the
desired results. For this reason, it is necessary to pre-test the questionnaire before it is used
in a full-scale survey, to identify any mistakes that need correcting.
The purpose of pre-testing the questionnaire is to determine:
a. whether the questions as they are worded will achieved the desired results
b. whether the questions have been placed in the best order
c. whether the questions are understood by all classes of respondents
d. whether additional or specifying questions are needed or whether some questions
should be eliminated
e. whether the instructions to interview are adequate
Usually a small number of respondents are selected for the pre-test. The
Respondents selected for the pilot survey should be broadly representative of the type of
respondent to be interviewed in the main survey. Respondents used for pre-test should no
longer be included in the main survey.
If the survey questionnaire has been subjected to a thorough pilot test, the final form
of the questions and questionnaire will have evolved into its final form. All that remains to be
done is the mechanical process of laying out and setting up the questionnaire in its final form.
QUANTITATIVE DATA ANALYSIS USING STATISTICS
Quantitative data analysis refers to the numerical representation and manipulation of
observations for the purpose of describing and explaining the phenomena that those
observations reflect. Quantitative data analysis is helpful in evaluation because it provides
quantifiable and easy to understand results.
Analysis of data involves a variety of descriptive and inferential statistics.
1. Descriptive Analyses – These analyses are called descriptive because they allow you
to summarize large amounts of information. Descriptive statistics include frequencies
(counts), percent’s, ranks, measures of central tendency (such as mean, median, and
mode), and measures of variability (such as range and standard deviation).
The most commonly used descriptive statistics are:
a. Mean – It indicates the average performance of a group on some measure of a
variable, and
b. Standard deviation – This indicates how spread out a set of scores is around the
mean, that is, whether the scores are relatively homogeneous or heterogeneous
c. around the mean.
 2. Inferential Analyses – After conducting descriptive analyses, you may want to conduct
more complex inferential analyses. These analyses include testing for significant
results.
The most commonly used inferential statistics are:
a. T-test – It is used to determine whether the means of two groups are statistically
different from one another;
b. Analysis of variance (ANOVA) – This is used to determine if there is significant
difference among the means of three or more groups; and
c. Chi square – It is used to compare group frequencies, or to see if an event occurs
more frequently in one group than another.
Linear Correlation
Linear correlation quantifies the strength of a linear relationship between two
numerical variables. When there is no correlation between two variables, there is no
tendency for the values of one quantity to increase or decrease with the values of the second
quantity. Linear correlation is one of the statistical techniques for bivariate analysis.
PRESENTATION AND INTERPRETATION OF DATA IN TABULAR AND
GRAPHICAL FORMS
It is likely that there will be occasions when you have numerical information that you
want to include in your work, for example, figures and other statistics from secondary sources
(such as books, journal articles or newspaper reports); the result of experiments; or data that
you have collected and analyzed as part of a research study.
The choice about whether to use text, tables or graphs requires careful consideration
if you are to ensure that your reader understands your argument and is not left struggling to
interpret data that are poorly presented or in an appropriate format.
Tables
Tables are used to present numerical data in a wide variety of publications from
newspapers, journals and textbooks to the sides of grocery packets. They are the format in
which most numerical data are initially stored and analyzed and are likely to be the means
you use to organize data collected during quantitative research. However, when writing up
your work you will have to make a decision about whether a table is the best way of
presenting the data, or if it would be easier to understand if you were to use a graph or chart.
Tables are an effective way of presenting data:
 a. When you wish to show how a single category of information varies when measured
a different point (in time or space).
b. When the dataset contains relatively few numbers. This is because it is very hard for
a reader to assimilate and interpret many numbers in table. In particular, avoid the
use of complex tables in talks and presentations when the audience will have a
relatively short time to take in the information and little or no opportunity to review
it at a later stage;
c. When you don’t wish the presence of one or two, very high or low numbers to detract
from the message contained in the rest of the dataset.

Figure 1. Sample Tabular Form

Table design
In order to ensure that your table is clear and easy to interpret there are a number of
design issues that need to be considered. These are listed below:
a. Since tables consist of rows and columns of information it is important to consider
how the data are arranged between the two. Most people find it easier to identify
patterns in numerical data by reading down a column rather than across a row. It is
easier to interpret the data if they arranged according to their magnitude so there is
numerical progression down the columns, although this may not always possible.
b. If there are several columns or categories of information a table can appear complex
and become hard to read. It also becomes more difficult to list the data by magnitude
 since the order that applies to one column may not be the same for others. In such
cases you need to decide which column contains the most important trend and this
should be used to structure the table. If the columns are equally important it is often
better to include two or more simple tables rather than using a single more complex
one.
c. Numbers in tables should be presented in their most simple format. This may mean
rounding up values to avoid the use of decimal places, stating the units (e.g. Php 4.6
million rather than Php 4,600,000) or using scientific notation (e.g. 6.315 X 10-2 rather
than 0.06315).
d. All tables should be presented with a title that contains enough detail that a reader
can understand the content without needing to consult the accompanying text. There
should also be information about the source of the data being used; this may be a
reference to a book or journal, or could indicate that the data are results from
experiment carried out on a particular date.
e. Where more than one table is being presented it is standard practice to give each one
a unique reference number, and in larger pieces of work, such as dissertations, a list
of tables with their page number is usually provided in addition to contents page.
f. The formatting of the table should not resemble a spreadsheet where each entry is
bounded by a box since this makes it difficult to read across rows or down columns.
However, the design of the table should help the reader interpret the data and so the
use of lines and/or bold text to separate headings from the body of data, or
highlighting/shading specific rows or may be effective. Avoid large gaps between
columns since this also makes it difficult to read along a row.
Graphs
Graphs are good means of describing, exploring or summarizing numerical data
because the use of a visual image can simplify complex information and help to highlight
patterns and trends in the data. They are a particularly effective way of presenting a large
amount of data but can also be used instead of a table to present smaller datasets. Using
graphs can help depict data and well-made graphs convey information quickly. There are
many different graph types to choose from and a critical issue is to ensure that the graph type
selected is the most appropriate for the data. Having done this, it is then essential to ensure
that the design and presentation of the graph help the reader or audience interpret the data.
Types of graphs
Different types of graphs are used for different situations. For this reason, it helps to
know a little bit about what the available graphs are. Many times, the kind of data is what
determines the appropriate graphs to use.
a. Bar charts – are one of the most commonly used types of graph and are used to
display and compare the number, frequency or other measure (e.g. mean) for
different discrete categories or groups. The graph is constructed such that the heights
or lengths of the different bars are proportional to the size of the category they
represent. Since the x-axis (the horizontal axis) represents the different categories it
has no scale. The y-axis (the vertical axis) does have a scale and this indicates the
units of measurement. The bars can be drawn either vertically or horizontally
depending upon the number of categories and length or complexity of the category
labels. The bars’ heights are scaled according to their values and the bars can be
compared to each other. Bar graphs can be drawn in a 3-dimensional way and
compiled for data comparison about the same thing or location. So that more
important categories are emphasized, bars in a bar graph are arranged in order of
frequency.
Figure 2. Sample Bar Chart represents the depth of the Great Lakes

b. Histograms – are a special form of bar chart where the data represent continuous
rather than discrete categories. However, because a continuous category may have
a large number of possible values the data are often grouped to reduce the number
 of data points. Unlike a bar chart, in a histogram both the x- and y-axes have a scale.
This means that it is the area of the bar that is proportional to the size of the category
represented and not just its height. Creating a histogram provides a visual
representation of data distribution. Histograms can display a large amount of data
and the frequency of the data values. The median and distribution of the data can be
determined by a histogram. In addition, it can show any outliers or gaps in the data.
Figure 3. Sample Histogram

c. Pie charts – are a visual way of displaying how the total data are distributed between
different categories. They are generally best for showing information grouped into a
small number of categories and are a graphical way of displaying data that might
otherwise be presented as a simple table. Sometimes called a circle graph, pie charts
represent the parts of a whole. Each section or slice of the pie is a data percentage.
From biggest to smallest, segments are arranged in a clockwise formation. This way,
the pie chart features easy-to-compare subjects presented in a neat, easy-to-
understand way.
 Figure 4. Sample Pie chart

d. Line graphs – are usually used to show time series data that is how one or more
variables vary over a continuous period of time. In a line graph the x-axis represents
the continuous variable (for example year or distance from the initial measurement)
while the y-axis has a scale and indicates the measurement. Several data series can
be plotted on the same line chart and this is particularly useful for analyzing and
comparing the trends in different datasets. Line graphs provide an excellent way to
map independent and dependent variables that are both quantitative. When both
variables are quantitative, the line segment that connects two points on the graph
expresses a slope, which can be interpreted visually relative to the slope of other lines
or expressed as a precise mathematical formula.
Figure 5. Sample Line Graph
 e. Scatter plots – are used to show the relationship between pairs of quantitative
measurements made for the same object or individual. Scatter plots are similar to
line graphs in that they start with mapping quantitative data points. The difference is
that with a scatter plot, the decision is made that the individual points should not be
connected directly together with a line but, instead express a trend. This trend can
be seen directly through the distribution of points or with the addition of a regression
line. A statistical tool used to mathematically express a trend in the data. Additionally,
with a scatter plot a mark, usually a dot or small circle, represents a single data point.
With one mark (point) for every data point a visual distribution of the data can be
seen. Depending on how tightly the points cluster together, you may be able to
discern a clear trend in the data
Figure 6. Sample Scatter Plot

DATA ANALYSIS AND INTERPRETATION

Data Analysis
Data analysis is at the heart of any scientific investigation. It involves working to
uncover patterns and trends in datasets.
In analyzing quantitative and qualitative data is often the topic of advanced research
and evaluation methods courses. However, there are certain basics which can help to make
sense of reams of data.
When analyzing data (whether from questionnaires, interviews, focus groups, or
whatever), always start from review of your research goals, i.e., the reason you undertook
the research in the first place. This will help you organize your data and focus your analysis.
For example, if you wanted to improve a program by identifying its strengths and weaknesses,
you can organize data into program strengths, weaknesses and suggestions to improve the
program. If you wanted to fully understand how your program works, you could organize data
in the chronological order in which customers or clients go through your program. If you are
conducting a performance improvement study, you can categorize data according to each
measure associated with each overall performance result, e.g., employee learning,
productivity and results. By publishing their data and the techniques they used to analyze and
interpret those data, scientists give the community the opportunity to both review the data
and use them in future research.
Basic analysis of "quantitative" information
1. Make copies of your data and store the master copy away. Use the copy for making
edits, cutting and pasting, etc.
2. Tabulate the information, i.e., add up the number of ratings, rankings, yes's, no's for
each question.
3. For ratings and rankings, consider computing a mean, or average, for each question.
For example, "For question #1, the average ranking was 2.4". This is more meaningful
than indicating, e.g., how many respondents ranked 1, 2, or 3.
4. Consider conveying the range of answers, e.g., 20 people ranked "1", 30 ranked "2",
and 20 people ranked "3".
Basic analysis of "qualitative" information
1. Read through all the data.
2. Organize comments into similar categories, e.g., concerns, suggestions, strengths,
weaknesses, similar experiences, program inputs, recommendations, outputs,
outcome indicators, etc.
3. Label the categories or themes, e.g., concerns, suggestions, etc.
4. Attempt to identify patterns, or associations and causal relationships in the themes,
e.g., all people who attended programs in the evening had similar concerns, most
people came from the same geographic area, most people were in the same salary
range, what processes or events respondents experience during the program, etc.
5. Keep all commentary for several years after completion in case needed for future
reference.
Discussion
Function: The function of the Discussion is to interpret your results in light of what was
already known about the subject of the investigation, and to explain our new understanding
of the problem after taking your results into consideration. The Discussion will always
connect to the Introduction by way of the question(s) or hypotheses you posed and the
literature you cited, but it does not simply repeat or rearrange the Introduction. Instead, it
tells how your study has moved us forward from the place you left us at the end of the
Introduction.
Fundamental questions to answer here include:
• Do your results provide answers to your testable hypotheses? If so, how do you
interpret your findings?
• Do your findings agree with what others have shown? If not, do they suggest an
alternative explanation or perhaps a unforeseen design flaw in your experiment (or
theirs?)
• Given your conclusions, what is our new understanding of the problem you
investigated and outlined in the Introduction?
• If warranted, what would be the next step in your study, e.g., what experiments
would you do next?
Style: Use the active voice whenever possible in this section. Watch out for wordy phrases;
be concise and make your points clearly. Use of the first person is okay, but too much use of
the first person may actually distract the reader from the main points.
Approach: Organize the Discussion to address each of the experiments or studies for which
you presented results; discuss each in the same sequence as presented in the Results,
providing your interpretation of what they mean in the larger context of the problem. Do
not waste entire sentences restating your results; if you need to remind the reader of the
result to be discussed, use "bridge sentences" that relate the result to the interpretation:
"The slow response of the lead-exposed neurons relative to controls suggests
that...[interpretation]".
You will necessarily make reference to the findings of others in order to support your
interpretations. Use subheadings, if need be, to help organize your presentation. Be wary of
mistaking the reiteration of a result for an interpretation, and make sure that no new
results are presented here that rightly belong in the results.
You must relate your work to the findings of other studies - including previous studies you
may have done and those of other investigators. As stated previously, you may find crucial
information in someone else's study that helps you interpret your own data, or perhaps you
will be able to reinterpret others' findings in light of yours. In either case you should discuss
reasons for similarities and differences between yours and others' findings. Consider how
the results of other studies may be combined with yours to derive a new or perhaps better
substantiated understanding of the problem. Be sure to state the conclusions that can be
drawn from your results in light of these considerations. You may also choose to briefly
mention further studies you would do to clarify your working hypotheses. Make sure
to reference any outside sources as shown in the Introduction section.
Do not introduce new results in the Discussion. Although you might occasionally
include in this section tables and figures which help explain something you are discussing,
they must not contain new data (from your study) that should have been presented earlier.
They might be flow diagrams, accumulation of data from the literature, or something that
shows how one type of data leads to or correlates with another, etc.
Data Interpretation
Data interpretation involves explaining those patterns and trends in datasets.
Scientists interpret data based on their background knowledge and experience; thus,
different scientists can interpret the same data in different ways. By publishing their data and
the techniques they used to analyze and interpret those data, scientists give the community
the opportunity to both review the data and use them in future research.
The analyzed data can then be interpreted and explained. In general, when scientists
interpret data, they attempt to explain the patterns and trends uncovered through analysis,
bringing all of their background knowledge, experience, and skills to bear on the question and
relating their data to existing scientific ideas. Given the personal nature of the knowledge
they draw upon, this step can be subjective, but that subjectivity is scrutinized through
the peer review process.
Interpreting information
1. Attempt to put the information in perspective, e.g., compare results to what you
expected, promised results; management or program staff; any common standards
 for your products or services; original goals (especially if you're conducting a program
evaluation); indications or measures of accomplishing outcomes or results (especially
if you're conducting an outcomes or performance evaluation); description of the
program's experiences, strengths, weaknesses, etc. (especially if you're conducting a
process evaluation).
2. Consider recommendations to help employees improve the program, product or
service; conclusions about program operations or meeting goals, etc.
3. Record conclusions and recommendations in a report, and associate interpretations
to justify your conclusions or recommendations.
Experimental design
The experimental design and analysis must be able to demonstrate that it is the
application of the product that is having the measured effect, rather than some other
variable. If a trial is poorly designed, even the most skilled statistician will have difficulty in
applying appropriate and meaningful statistical analysis to prove that the product works as
claimed.
Completely Randomized Design
The completely randomized design is the simplest experimental design. In this design,
treatments are replicated but not blocked, which means that the treatments are assigned to
plots in a completely random manner (as in the left side of figure 7). This design is
appropriate if the entire test area is homogeneous (uniform in every way that can influence
the results). Unfortunately, it is rare that you can ever be confident of a test site's uniformity,
so a completely randomized design is rarely used in field tests. The completely randomized
design is used more commonly in greenhouse tests, though blocking is often useful even in
the more controlled environment of a greenhouse.
Figure 7. The shaded area represents an area of the field that is different from the unshaded
area. Treatments (A, B, and C) are replicated but not blocked in the field on the left. On the
right, treatments are replicated and blocked; each block contains one plot of each
treatment.

Randomized Complete Block Design

The randomized complete block design is the most commonly used design in
agricultural field research. In this design, treatments are both replicated and blocked, which
means that plots are arranged into blocks and then treatments are assigned to plots within a
block in a random manner (as in the right side of figure 7). This design is most effective if you
can identify the patterns of non-uniformity in a field such as changing soil types, drainage
patterns, fertility gradients, direction of insect migration into a field, etc. If you cannot identify
the potential sources of variation, you should still use this design for field research but make
your blocks as square as possible. This usually will keep plots within a block as uniform as
possible even if you cannot predict the variation among plots.
Blocking refers to physically grouping treatments together in an experiment to
minimize unexplained variation in the data you collect (referred to as experimental error).
This allows the statistical analysis to identify treatment differences that would otherwise be
obscured by too much unexplained variation in the experiment. Variation in an experiment
can be divided into two types: variation for which you can account in the statistical analysis
and variation that is unexplained. The goal in blocking is to allow you to measure the variation
among blocks and then remove that variation from the statistical comparison of treatment
means. If you can anticipate causes of variation, you can block the treatments to minimize
variation within each block and remove some variation from the statistical analysis. The
mathematics of how blocking allows you to reduce unexplained variation is beyond the scope
of this bulletin.
In the most common experimental designs, a block will contain one plot of each
treatment in the experiment. If an experiment has five treatments, then each block will
contain five plots, with each plot receiving a different treatment. When a block contains one
plot of each treatment, then each block represents one replication of each treatment. For this
reason, blocks are frequently referred to as "replications" or "reps," but the concept of
blocking should not be confused with the concept of replication; replication and blocking
serve different purposes. In agricultural research, field plots are almost always blocked even
when no obvious differences are present in the field. It is much better to block when you did
not really need to than not to block when you should have blocked.

Figure 8. An easy way to arrange blocks is to put them side by side across the field. Letters
represent different treatments.

Blocking is a very powerful tool that is most effective if you can anticipate sources of
variation before you begin an experiment. For example, in an herbicide trial, one side of a
field may have a history of more severe weed problems. If you just scattered your treatments
randomly through the field, a lot of the variation in the data you collected could be due to the
increased weed pressure on one side of the field. Such variation would make it difficult to
determine how well each treatment worked. Because you know one side of the field will have
more weeds, you can remove that source of variation from the statistical analysis by blocking
and improve your chances of identifying differences among treatments.
 The process of blocking follows a logical sequence. First, you determine that there is
something (weeds, drainage, sun/shadow, water, soil type, etc.) that is not uniform
throughout the experimental area (field, greenhouse, etc.) that may influence whatever you
are measuring (yield, plant height, etc.). Then you can arrange your treatments into blocks so
that the area within each block is as uniform as possible (see figure 7). Though the area within
a block should be relatively uniform, there may be large differences among the blocks, but
that is what makes blocking effective. Your goal is to maximize the differences among blocks
while minimizing the differences within a block.
The shape of the blocks is not important as long as the plots within a block are as
uniform as possible. Ideally, the only differences among plots within a block should be due to
the treatments. Blocks in field experiments are usually square or rectangular, but they may
be any shape. Blocks in the same experiment do not have to be the same shape; the shape of
individual blocks will be determined by variation in the field that you are trying to minimize.
If you are not sure what shape your blocks should be, square or nearly square blocks are
usually a safe choice.
Blocks may be arranged through the field in many ways. If the field is wide enough, an
easy way to arrange blocks is to place them side-by-side all the way down the field (see figure
8). But blocks do not have to be contiguous and may be scattered through the field in any way
that is convenient for you.
Factorial Arrangement of Treatments
A factorial arrangement of treatments is not an experimental design, though you will
often hear it referred to as a factorial design or a factorial experiment. A factorial
arrangement of treatments means that the experiment is testing two or more factors at the
same time, and that the experiment includes all combinations of all factors. The term "factor"
is used to describe a group of treatments that have something in common. Fungicides,
sources of nitrogen, or corn hybrids could be considered factors in an experiment. Factors
may be defined broadly or narrowly in different experiments. All herbicides may be grouped
as a factor in one experiment, but pre-plant and post-plant herbicides may be treated as
separate factors in another experiment. A single-factor experiment tests one factor at a time;
a two-factor experiment tests two factors at once.
Figure 9. A 2x5 factorial arrangement of treatments in a randomized complete block design
(above) and in a split-plot design (below). A and B represent two levels of one factor, and
the numbers (1-5) represent five levels of a second factor. The combinations (e.g., 4A, 5B,
etc.) denote individual treatment combinations. Either experimental design could be used,
but the randomized complete block design is preferred unless the split-plot design is
required by some limitation on randomization.
 Most simple on-farm experiments are single-factor experiments (in a Completely
Randomized or Randomized Complete Block design) and compare things such as crop
varieties or herbicides, but it is sometimes useful to test two or more factors at once. For
example, a two-factor experiment would allow you to compare the yields five corn hybrids at
three planting dates.
This accomplishes three things at once:
➢ It allows you to compare the corn hybrids with each other.
➢ It allows you to evaluate the effect of planting date.
➢ It allows you to determine if varying the planting date changes the relative
performance of the hybrids (e.g. one hybrid may only perform well if planted early).
The first two could be done in separate single-factor experiments, but the third can
only be achieved by having both factors in a single experiment. This becomes especially
important if one factor can have a significant influence on the effect of the other factor. For
example, you might test soybean varieties as one factor and nematicides as another factor. If
a few varieties have good nematode resistance but others do not, they may appear equally
good when effective nematicides are used but varieties with resistance would appear much
better when nematicides are not used. In cases like this, the effect of one factor (variety) is
strongly influenced by the other factor (nematicide). When one factor influences the effect
of the other factor, there is said to be a significant interaction between the two factors. It can
be very important to know if there is an interaction between factors, because if there is an
interaction, you can make predictions or recommendations based on the results of single-
factor experiments only when all other factors are at the same levels they were at in the
experiment. If you change some factor not included in the experiment, the results from your
single-factor experiment may no longer be valid.
With a factorial arrangement of treatments, all values (or levels) of each factor must
be paired with all levels of the other factors. If you have two nematicides and five soybean
varieties, then your treatment list must include each variety with each nematicide for a total
of 10 treatments. This would be referred to as a "two by five factorial" to denote how many
factors were present in the experiment and how many levels of each factor were used. The
number of treatments increases quickly when you add more levels for a factor (if you used
three nematicides instead of two, you would have 15 treatments instead of 10), so choose
your levels carefully or the experiment can get too large to manage.
 A factorial arrangement of treatments can be a very powerful tool, but because the
number of treatments can get very large it is best used when some reason exists to believe
that the factors may influence each other and have a significant interaction. If there is no
suspicion that the factors may influence each other, it is frequently easier and more thorough
to test the factors in separate experiments. A factorial arrangement of treatments can be
used with a completely randomized experimental design or a randomized complete block
design. The top half of figure 4 shows a factorial arrangement of treatments in a randomized
complete block design.
Split-Plot Experimental Design
A split-plot experimental design is a special design that is sometimes used with
factorial arrangements of treatments. This design usually is used when an experiment has at
least two factors and some constraint prevent you from randomizing the treatments into a
randomized complete block design. Such a constraint may be based on equipment limitations
or on biological considerations. For example, the equipment you have may make it difficult to
put out a soil fumigant in randomized complete blocks, but you may be able to put out the
fumigant so that all treatments within a block that get the fumigant will be clustered together
rather than scattered throughout the block. You can use a split-plot experimental design to
work around this limitation as long as you are able to randomize the other factors. There are
other situations when this design is appropriate, but a constraint on randomization is the
most likely to occur.
Suppose you want to test the effect of five fungicides to control Cylindrocladium Black
Rot on two varieties of peanut. In this test, you would have a 2x5 factorial arrangement of
treatments: The two factors would be varieties (2 levels of this factor) and fungicides (5 levels
of this factor). Because a factorial arrangement of treatments is not an experimental design,
you still have to select an experimental design that best meets your needs. If you are able to
randomize varieties and fungicides within a block, then you should pick a randomized
complete block design. If there is some reason why you cannot completely randomize the
treatments within each block, then you may be able to use a split-plot design to work around
that limitation. For example, you may have a six-row planter but only enough space in the
field to put out four-row plots. To resolve this dilemma, you could plant all of the plots that
have the same peanut variety together within a block and then randomize the five fungicide
treatments within each peanut variety.
 In split-plot designs, the terms "whole plots" and "sub-plots" refer to the plots into
which the factors are randomized. As the names imply, whole plots are subdivided into
subplots. In figure 9, a whole plot would be the areas designated with A or B, and the subplots,
the subdivisions within the whole plots, are designated 1, 2, 3, 4, or 5. In this example, A and
B could represent two varieties (two levels of one factor) and the numbers could represent
different fungicides (five levels of a second factor). Each whole plot serves as a block for the
subplot treatments.
To assign treatments in a split-plot design, start by identifying where each block will
be. Then randomize the whole plot treatments within each block. The whole plot treatments
will be the treatment that you are unable to randomize into a randomized complete block
design. The subplot treatments can then be randomized within each whole plot treatment
(see figure 9).
Experimental Analysis
The type of analysis to be conducted depends on the purpose and design of the trial
and the type of observations made. Statistical analysis is not required in all cases; nor is it
appropriate in certain situations. However, when a comparison between two products or
between one product and no treatment is required, statistical analysis must be provided to
support the interpretation of the data. Novel statistical analysis submitted in support of
experimental data should be accompanied by the raw data and the published literature that
references the statistical technique. This guideline cannot describe all analytical approaches
for all trial designs, but aims to provide some principles of analysis to assist applicants. If you
are not confident of your knowledge in this area, it is highly advisable to seek the assistance
of a competent statistician before starting trial work.
Typically, it is the variable that determines which broad type of analysis is required
(that is, parametric or non-parametric). If the variable is quantitative (binary, binomial,
discrete or continuous), parametric statistical methods should be used, such as analysis of
variance or linear or logistic regression. If the variable is qualitative (nominal and ordinal
methods, such as ranking or scoring), non-parametric methods are required.
Before conducting a parametric analysis of variance, three assumptions should be met to
ensure that the analysis is valid:
➢ additivity of effects
➢ homogeneity of variance
 ➢ normality of the error.
If these three assumptions cannot be met, non-parametric methods may be preferred.
Parametric tests
Additivity requires that the sources of variability (e.g. treatments) are independent of
each other. Independence results in an additive (e.g. multiplicative or logarithmic) effect on
the response variable (e.g. pest population). The more variables interact with each other, the
greater the chance that the observed response is not the result of the individual treatments
may invalidate the observed results. Sometimes, effect results are not on a natural scale and
must be transformed to different scales (for example, probit or logit) to meet additivity
requirements. Methods to test additivity are available (such as Tukey’s test of additivity).
Homogeneity of variance requires that all the populations tested contain the same
level of variability. The less homogeneity between variances of populations being compared,
the less likely it is that a parametric method will be able to accurately produce a significant
result. There are many tests used to test homogeneity of variance, each with advantages and
disadvantages.
Normality requires the distribution of errors (variance around the mean) to be normal.
Normal distribution is important because the further the distributions are from normal, the
less validity any analysis of variance assessments will have, as there is a greater chance that a
significant result will be false (and vice versa). Standard tests and graphical displays are
available to demonstrate normality.
Analysis of variance
When reporting the results of an analysis of variance (ANOVA), you should present a
table of means of each of the treatments, along with the standard error or confidence interval
(the variability around the mean). Presenting means with the variability of results can
overcome the difficulty of explaining statistically equivalent results when differences
between means are large (and vice versa).
Formal statistical tests, often as F-tests, are usually also performed to demonstrate
any significant results between treatments. Typically, study reports present an analysis to
compare all treatment means against each other. In considering the original objective of the
trial, this may not be necessary and may confuse the analysis and interpretation. For example,
not all treatments need be compared against each other, especially if the comparison of
interest is only a limited set of treatments, such as the new product versus the industry
standard at proposed label rates. If the trial is designed for this purpose, these matters should
be considered at the trial design stage (for example, t-tests may be appropriate). You should
consult appropriate texts and professional statisticians if you are unsure of the most
appropriate test or procedure.
Non-parametric tests
Non-parametric methods may be required and are preferred when the data are
qualitative rather than quantitative or the three assumptions described above cannot be met.
However, non-parametric methods should be used with caution when analyzing small data
sets. There are a number of different non-parametric methods, many of which are suitable
only in certain situations. You may wish to refer to the EPPO’s Design and analysis of efficacy
evaluation trials (PP 1/152(4))(link is external) for references describing which test is relevant
to a particular type of data set.
Trial series
You may need to conduct separate but closely similar trials at different locations
and/or at different times. The series of trials can be analyzed together in certain
circumstances (for example, if they have the same methods, external impacting factors and
pest abundance and similar standard error) and for particular reasons (for example, to
estimate treatment effects over sites and years or to test potential confounding factors). Such
an analysis should not be conducted unless it has been planned for at the trial design stage
so that all requirements can be met. See EPPO guideline PP 1/152(4)(link is external) for more
details.
Interpretation of Experimental Data
This section answers the question, “So what?” in relation to the results of the study.
What do the results of the study mean? This part is, perhaps, the most critical aspect of the
research report. It is often the most difficult to write because it is the least structured. This
section demands perceptiveness and creativity from the researcher.
How do we interpret the result(s) of our study?
1. Tie up the results of the study in both theory and application by pulling together the:
a. conceptual/theoretical framework
b. review of literature: and
c. the study’s potential significance for application
2. Examine, summarize, interpret and justify the results; the draw inferences. Consider
the following:
a. Conclude or summarize
• This technique enables the reader to get the total picture of the findings
in summarized form, and helps orient the reader to the discussion that follows.
b. Interpret
• Questions on the meaning of the findings, the methodology, the
unexpected
• Results and the limitations and shortcomings of the study should be
answered and interpreted.
c. Integrate
• This is an attempt to put the pieces together.
• Often, the results of a study are disparate and do not seem to “hang
together”. In the discussion, attempt to bring the findings together to
extract meaning and principles.
d. Theorize
When the study includes a number of related findings, it occasionally becomes
possible to theorize.
• Integrate your findings into a principle;
• Integrate a theory into your findings; and
• Use these findings to formulate and original theory
e. Recommend or apply alternatives
REVIEW QUESTIONS FOR DISCUSSIONS
Direction: Answer the following questions:
1. What are the advantages and disadvantages of using survey instruments?
2. When preparing a questionnaire what types of questions to be avoided.
3. When are tables effective in presenting data? Specify.
4. What is a good table design? Discuss briefly.
5. What are the different types of graphs for presenting data? Describe each in detail.
6. What are the most commonly used descriptive and inferential statistics in the analysis
of data? Discuss briefly.
7. What are the different experimental designs used to analyze the results? Discuss
briefly.
8. How do we interpret the result(s) of our study?