The purpose of this experiment was to examine whether font type and its consistency at encoding and retrieval stages had any influence on the encoding specificity principle. Based on prior research which resulted in higher performances in memory recall due to retrieval cues, it is predicted that there will be a higher performance of recognition for participants who have consistency of font between retrieval and encoding stages and no variability at the encoding stage. The experiment contained two independent variables with two between-subjects levels (variability of encoding font, consistency of font between retrieval and encoding). Participants included 60 Binghamton University students that were randomly placed into one of four experimental groups. Each participant was given a 30 word list to memorize. Then a 60 word list was distributed and participants were instructed to circle the words they could recognize from the encoding list. No significant results were found, meaning variability of font at encoding and consistency of font between retrieval and encoding stages had no significant difference in participants’ performance. Our lack of significant results could be due to the high skill level and intelligence of the participants used in the study.Keywords: Encoding Specificity, Recognition, Retrieval CuesDoes Font Have an Effect on the Encoding Specificity Principle and Recognition?The encoding specificity hypothesis stated by Tulving and Thomson (1971) suggests the probability of remembering the original encoding of information will be greater if the conditions at encoding are the same conditions at retrieval (Tulving & Thomson, 1971).  Moscovitch and Craik (1976) tested the ability to recall encoded words after either being asked encoding questions (retrieval cues) during recall or given no retrieval cues as aids.  Retrieval cues are stimuli that help people retrieve certain memories. When new memories are formed, individuals include information about the situation that acts as a trigger to find and access that saved memory.  Moscovitch and Craik found that recall was significantly higher for participants that were given retrieval cues as aids. The presence of the encoding questions at both the time of encoding and retrieval increased retention in memory and facilitated the retrieval of the encoding information (Moscovitch & Craik, 1976). This study provides evidence that the encoding specificity principle exists and that retrieval cues are an effective factor in retention of memory.Researchers began to focus their studies to find conditions that resulted in the presence of encoding specificity.  Eich did a study based on the previous work mentioned by Moscovitch and Craik where she utilized the use of retrieval cues, but focused on categories and rhyming words. Examples would be questions like “Is the word a fruit?” or “Does the word rhyme with hair?”. The answers to those questions would either be positive (yes) or negative (no). Results from this study found that category-positive responses were significantly higher in recall than rhyme-positive responses and all negative responses (Eich, 1985). In other words, categorizing the terms resulted in the highest performances of recall. This study supported the positive use of retrieval cues in memory retention.  It also provides evidence for the encoding specificity principle based on its use of encoding questions at the encoding and retrieval stage which led to higher performances in recall. It was questioned whether these findings in recall would correlate to recognition.  In the past, it was believed that less information was needed to pass a recognition task than a recall task because recognition is the ability to identify a familiar piece of information while recall is the ability to fully remember a specific piece of information.  For example, if you are trying to remember the name of your first grade teacher,  knowing the answer on the spot would be recall. If you write down all the names of the teachers you ever had and selected the answer from that list, that would be considered recognition. Experimenters considered that  partial learning may be enough to distinguish between correct and incorrect answers to a recognition task (Postman, 1963).   Tversky (1973) conducted an experiment on recognition and recall using picture stimuli and found results that went against the notion that less information is needed for recognition.  Both recognition and recall had similar numbers in performance in this study. Although these two processes encode, store and retrieve information differently, they both are processes of memory and had similar results in this particular study.In our study we focused on the encoding specificity principle with the use of font as a retrieval cue. With the knowledge that retrieval cues bolster performance of recall, we hypothesized that having consistent font at encoding and retrieval stages would increase the ability to recognize words. We also hypothesized that if there was no variability on the encoding list then performance in recognition would be higher due to less distraction of contrasting fonts.  We believed that previous findings found for the process of recall would carry-over to recognition to a certain degree based on the study done by Tversky. Our goal was to determine whether these conditions of font fall into the encoding specificity principle and if font is an effective retrieval cue. MethodDesign We conducted a 2 (encoding font variability: all one font, different fonts) x 2 (retrieval font consistency with encoding: same font as encoding font, different fonts than encoding font) between-subjects experiment.  The dependent variable of the study was the number of words correctly recognized.Participants Participants were drawn from a level 300 psychology class at Binghamton University (N=60; 33 female, 27 male). The mean age of the sample was 20.6 years old with a standard deviation of 0.995.  Each participant was assumed literate based on their enrollment in the university.  Participants were not compensated for taking part in the study.Materials Participants were randomly assigned to one of four experimental groups in a between-subjects design. Each participant received the experimental materials consisting of a sheet with an encoding list on the front and a retrieval list on the back. Four different sheets were used representing each cell of the 2 x 2 design, specifically: non-variable font encoding list/consistency of font between retrieval and encoding, non-variable font encoding list/inconsistency of font between retrieval and encoding,variable font encoding list/consistency of fonts between retrieval and encoding, variable encoding list/inconsistency of font between retrieval and encoding (see Appendix for all lists used in the study). Each encoding list consisted of the same 30 words that were randomly selected from a word generator.  Each recognition list consisted of those same 30 words on the encoding side plus another 30 words which were randomly selected as well  (total of 60 words for recognition list). Words were spaced slightly closer on the recognition side to fit the words on one side of a sheet of paper.  Word placement on the page was also changed at random between encoding and retrieval sides to make the recognition task more difficult.Non-Variable Consistent. This experimental group received the sheet which consisted of non-variable font on encoding side and had consistency between retrieval and encoding font.  In other words, it had one font on the encoding side and that same font on the recognition side.  This sheet used the Arial font in size 18. Non-Variable Inconsistent. This experimental group received the lists with non-variable font on the encoding side and inconsistent fonts between retrieval and encoding side.  In other words, it had one font on the encoding side and three different fonts on the recognition side. The font on the encoding side was Arial in size 18.  Three fonts were used on the recognition side and were chosen at random. These fonts were Times New Roman in size 18, Courier New in size 18 and Bad Script in size 18.  The words were assigned one of the three fonts at random. All fonts were used an equal amount of times on the side. That is to say, Times New Roman font was used for 20 words, Courier New was used for 20 words and Bad Script was used for 20 words. Variable Consistent.  This experimental group  received the lists with variable fonts on the encoding side and consistent fonts between retrieval and encoding side. In other words, there were different fonts on the encoding side and those same fonts on the recognition side.  These fonts were Times New Roman in size 18, Courier New in size 18 and Bad Script in size 18.  To specify, if a word was in Bad Script on the encoding side, then it was in Bad Script on the retrieval side. Variable Inconsistent.  This experimental group  received the lists with variable fonts on the encoding side and inconsistent fonts between retrieval and encoding sides.  In other words, there were different fonts on the encoding side and contrasting different fonts on the recognition side. The encoding side fonts were Times New Roman in size 18, Courier New in size 18 and Bad Script in size 18. The recognition fonts were Impact in size 18, Pacifico in size 18 and Syncopate in size 12.  Syncopate was put in size 12 instead of size 18 because the words showed up bigger on the sheet as compared to the other words. We did this to keep consistency of how the words looked on the page.Procedure The procedures for each of the four experimental groups were identical to each other.  The experimental materials were handed out and participants were instructed not to look down when they were received. Participants were allotted 45 seconds to look at the encoding list and memorize the words to the best of their ability for a recognition task after. Participants were then instructed they had 60 seconds to circle the words they recognized on the retrieval list. Participants were told to not circle aimlessly.  Afterwards, participants were asked to self report their gender and age on the bottom of the provided paper. It should be noted that most participants did not use the full 60 seconds, and handed in their paper when they could not recall any more words. Results A two-way ANOVA was run on the program RStudio  to determine if there was a  significant effect for variability of the encoding font alone, consistency of retrieval font with encoding font alone, and an interaction between the two variables.  All statistical comparisons used the significance level of p < .05. There was no significant effect of randomness of encoding font alone, p = .299. Performance of word recognition with all the same encoding font (M = 13.96, SD = 5.28) was not significantly different from words being in three different fonts (M =13.76, SD = 5.00). There was also no significant effect of consistency of retrieval font with encoding font alone, p = .881. Performance of word recognition with consistency between the retrieval and encoding fonts (M = 14.57, SD = 5.07) was not substantially different from words that were inconsistent in font between retrieval and encoding (M = 13.17, SD = 5.11). There was no significant effect of the interaction, p = .619. No post-hoc tests were done because no significant results were found.  A summary of recognition across the four experimental groups is shown in Figure 1. Discussion Due to the previous studies done on encoding specificity, we hypothesized performance on the recognition task would be significantly higher for participants that had no variability on the encoding list. We also hypothesized that performances would be higher for the recognition task for participants that had consistent fonts between retrieval and encoding stages. The results of this experiment did not support our hypotheses. In our study, the sample of participants used may have mattered because of their assumed degree of intelligence and skill level of recognition.  The fact each participant was an enrolled college student at a premier university could possibly be a ceiling effect in the study.  Perhaps if our sample was not so skilled, the distinctiveness of the font would've showed encoding specificity.  Location of distribution and participation in the study also could be a factor leading to our non-statistically significant results.  Data was collected in a loud, unorganized room with other experiments being conducted on different topics.  Also, students were taking part in multiple studies in a short period of time which could have led to a crossover of words and higher chance of mistakes.. It was a difficult place to focus and give all of one's full attention to the task in front of them. Degree of difficulty of the task could be another factor leading to our obtained results.  Changes could be made to increase the difficulty of our task and potentially get significant results if tested again.  The time period of encoding could be made smaller or the fonts could be made more difficult to read. Some other simple changes could be made to make the experiment go smoother. A cover sheet could have been made to block participants from looking down before the time had started.  The two lists could also be put on different sheets of paper and put in order so there would be no confusion of which list was the encoding list and which list was the recognition list. It was believed that if our sample size was bigger we would have had significant results.  A sampling procedure was done on the program Rstudio to increase our data four times than what it was to see if this would be true. In other words, our 60 participants were increased to 240 participants. This showed us a significant result of consistency between retrieval and encoding font.  This significant result would directly correlate with encoding specificity. For an exploratory analysis, we could say that if we had more participants, we would find the effect of encoding specificity come out in our results.Further attempts should be made to test whether encoding specificity truly exists for font studies.  Experimenters should redo this experiment with a larger sample of more moderately skilled individuals. We would suggest conducting the experiment in a more quiet and calm location. The idea of retrieval cues with font could assist in many everyday situations and actions. One example is a college student studying.  If it was known that learning in a specific font and studying in that same font positively affected memory retention, students could change their study habits to learn more easily and obtain better grades.  Another example could be an Alzheimer's patient trying to train their brain and increase their memory threshold. Although our results were not statistically significant, it is known that encoding specificity exists.  There is still a challenge for future experimenters to find more evidence on whether font is an effective retrieval cue.